diff options
| author | Pavel Zakharov <[email protected]> | 2016-07-22 10:39:36 -0400 |
|---|---|---|
| committer | Brian Behlendorf <[email protected]> | 2018-05-08 21:35:27 -0700 |
| commit | 6cb8e5306d9696d155ae7a808f56c4e46d69b64c (patch) | |
| tree | c5c1f331a6341281fd3e9b7310e1618eab2c005c | |
| parent | afd2f7b7117ff8bf23afa70ecae86ec0c1a1461e (diff) | |
OpenZFS 9075 - Improve ZFS pool import/load process and corrupted pool recovery
Some work has been done lately to improve the debugability of the ZFS pool
load (and import) process. This includes:
7638 Refactor spa_load_impl into several functions
8961 SPA load/import should tell us why it failed
7277 zdb should be able to print zfs_dbgmsg's
To iterate on top of that, there's a few changes that were made to make the
import process more resilient and crash free. One of the first tasks during the
pool load process is to parse a config provided from userland that describes
what devices the pool is composed of. A vdev tree is generated from that config,
and then all the vdevs are opened.
The Meta Object Set (MOS) of the pool is accessed, and several metadata objects
that are necessary to load the pool are read. The exact configuration of the
pool is also stored inside the MOS. Since the configuration provided from
userland is external and might not accurately describe the vdev tree
of the pool at the txg that is being loaded, it cannot be relied upon to safely
operate the pool. For that reason, the configuration in the MOS is read early
on. In the past, the two configurations were compared together and if there was
a mismatch then the load process was aborted and an error was returned.
The latter was a good way to ensure a pool does not get corrupted, however it
made the pool load process needlessly fragile in cases where the vdev
configuration changed or the userland configuration was outdated. Since the MOS
is stored in 3 copies, the configuration provided by userland doesn't have to be
perfect in order to read its contents. Hence, a new approach has been adopted:
The pool is first opened with the untrusted userland configuration just so that
the real configuration can be read from the MOS. The trusted MOS configuration
is then used to generate a new vdev tree and the pool is re-opened.
When the pool is opened with an untrusted configuration, writes are disabled
to avoid accidentally damaging it. During reads, some sanity checks are
performed on block pointers to see if each DVA points to a known vdev;
when the configuration is untrusted, instead of panicking the system if those
checks fail we simply avoid issuing reads to the invalid DVAs.
This new two-step pool load process now allows rewinding pools accross
vdev tree changes such as device replacement, addition, etc. Loading a pool
from an external config file in a clustering environment also becomes much
safer now since the pool will import even if the config is outdated and didn't,
for instance, register a recent device addition.
With this code in place, it became relatively easy to implement a
long-sought-after feature: the ability to import a pool with missing top level
(i.e. non-redundant) devices. Note that since this almost guarantees some loss
of data, this feature is for now restricted to a read-only import.
Porting notes (ZTS):
* Fix 'make dist' target in zpool_import
* The maximum path length allowed by tar is 99 characters. Several
of the new test cases exceeded this limit resulting in them not
being included in the tarball. Shorten the names slightly.
* Set/get tunables using accessor functions.
* Get last synced txg via the "zfs_txg_history" mechanism.
* Clear zinject handlers in cleanup for import_cache_device_replaced
and import_rewind_device_replaced in order that the zpool can be
exported if there is an error.
* Increase FILESIZE to 8G in zfs-test.sh to allow for a larger
ext4 file system to be created on ZFS_DISK2. Also, there's
no need to partition ZFS_DISK2 at all. The partitioning had
already been disabled for multipath devices. Among other things,
the partitioning steals some space from the ext4 file system,
makes it difficult to accurately calculate the paramters to
parted and can make some of the tests fail.
* Increase FS_SIZE and FILE_SIZE in the zpool_import test
configuration now that FILESIZE is larger.
* Write more data in order that device evacuation take lonnger in
a couple tests.
* Use mkdir -p to avoid errors when the directory already exists.
* Remove use of sudo in import_rewind_config_changed.
Authored by: Pavel Zakharov <[email protected]>
Reviewed by: George Wilson <[email protected]>
Reviewed by: Matthew Ahrens <[email protected]>
Reviewed by: Andrew Stormont <[email protected]>
Approved by: Hans Rosenfeld <[email protected]>
Ported-by: Tim Chase <[email protected]>
Signed-off-by: Tim Chase <[email protected]>
OpenZFS-issue: https://illumos.org/issues/9075
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/619c0123
Closes #7459
35 files changed, 2841 insertions, 553 deletions
diff --git a/cmd/zpool/zpool_main.c b/cmd/zpool/zpool_main.c index 453fb2131..d44589303 100644 --- a/cmd/zpool/zpool_main.c +++ b/cmd/zpool/zpool_main.c @@ -1820,6 +1820,10 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name, (void) printf(gettext("currently in use")); break; + case VDEV_AUX_CHILDREN_OFFLINE: + (void) printf(gettext("all children offline")); + break; + default: (void) printf(gettext("corrupted data")); break; @@ -1919,6 +1923,10 @@ print_import_config(status_cbdata_t *cb, const char *name, nvlist_t *nv, (void) printf(gettext("currently in use")); break; + case VDEV_AUX_CHILDREN_OFFLINE: + (void) printf(gettext("all children offline")); + break; + default: (void) printf(gettext("corrupted data")); break; @@ -2752,6 +2760,7 @@ zpool_do_import(int argc, char **argv) idata.guid = searchguid; idata.cachefile = cachefile; idata.scan = do_scan; + idata.policy = policy; pools = zpool_search_import(g_zfs, &idata); diff --git a/include/libzfs.h b/include/libzfs.h index cbaaa13a2..45eb5c904 100644 --- a/include/libzfs.h +++ b/include/libzfs.h @@ -413,6 +413,7 @@ typedef struct importargs { int unique : 1; /* does 'poolname' already exist? */ int exists : 1; /* set on return if pool already exists */ int scan : 1; /* prefer scanning to libblkid cache */ + nvlist_t *policy; /* rewind policy (rewind txg, etc.) */ } importargs_t; extern nvlist_t *zpool_search_import(libzfs_handle_t *, importargs_t *); diff --git a/include/sys/fs/zfs.h b/include/sys/fs/zfs.h index de3b729eb..fa4eb2721 100644 --- a/include/sys/fs/zfs.h +++ b/include/sys/fs/zfs.h @@ -704,6 +704,7 @@ typedef struct zpool_rewind_policy { #define ZPOOL_CONFIG_VDEV_TOP_ZAP "com.delphix:vdev_zap_top" #define ZPOOL_CONFIG_VDEV_LEAF_ZAP "com.delphix:vdev_zap_leaf" #define ZPOOL_CONFIG_HAS_PER_VDEV_ZAPS "com.delphix:has_per_vdev_zaps" +#define ZPOOL_CONFIG_CACHEFILE "cachefile" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_STATE "mmp_state" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_TXG "mmp_txg" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_HOSTNAME "mmp_hostname" /* not stored on disk */ @@ -811,6 +812,7 @@ typedef enum vdev_aux { VDEV_AUX_BAD_ASHIFT, /* vdev ashift is invalid */ VDEV_AUX_EXTERNAL_PERSIST, /* persistent forced fault */ VDEV_AUX_ACTIVE, /* vdev active on a different host */ + VDEV_AUX_CHILDREN_OFFLINE, /* all children are offline */ } vdev_aux_t; /* diff --git a/include/sys/spa.h b/include/sys/spa.h index 1172468ad..8a3938e86 100644 --- a/include/sys/spa.h +++ b/include/sys/spa.h @@ -410,6 +410,7 @@ typedef enum bp_embedded_type { #define SPA_BLKPTRSHIFT 7 /* blkptr_t is 128 bytes */ #define SPA_DVAS_PER_BP 3 /* Number of DVAs in a bp */ +#define SPA_SYNC_MIN_VDEVS 3 /* min vdevs to update during sync */ /* * A block is a hole when it has either 1) never been written to, or @@ -1015,11 +1016,16 @@ extern boolean_t spa_has_pending_synctask(spa_t *spa); extern int spa_maxblocksize(spa_t *spa); extern int spa_maxdnodesize(spa_t *spa); extern void zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp); +extern boolean_t zfs_dva_valid(spa_t *spa, const dva_t *dva, + const blkptr_t *bp); typedef void (*spa_remap_cb_t)(uint64_t vdev, uint64_t offset, uint64_t size, void *arg); extern boolean_t spa_remap_blkptr(spa_t *spa, blkptr_t *bp, spa_remap_cb_t callback, void *arg); extern uint64_t spa_get_last_removal_txg(spa_t *spa); +extern boolean_t spa_trust_config(spa_t *spa); +extern uint64_t spa_missing_tvds_allowed(spa_t *spa); +extern void spa_set_missing_tvds(spa_t *spa, uint64_t missing); extern boolean_t spa_multihost(spa_t *spa); extern unsigned long spa_get_hostid(void); diff --git a/include/sys/spa_impl.h b/include/sys/spa_impl.h index 90d929268..cd214c29c 100644 --- a/include/sys/spa_impl.h +++ b/include/sys/spa_impl.h @@ -184,6 +184,15 @@ typedef enum spa_all_vdev_zap_action { AVZ_ACTION_INITIALIZE } spa_avz_action_t; +typedef enum spa_config_source { + SPA_CONFIG_SRC_NONE = 0, + SPA_CONFIG_SRC_SCAN, /* scan of path (default: /dev/dsk) */ + SPA_CONFIG_SRC_CACHEFILE, /* any cachefile */ + SPA_CONFIG_SRC_TRYIMPORT, /* returned from call to tryimport */ + SPA_CONFIG_SRC_SPLIT, /* new pool in a pool split */ + SPA_CONFIG_SRC_MOS /* MOS, but not always from right txg */ +} spa_config_source_t; + struct spa { /* * Fields protected by spa_namespace_lock. @@ -202,6 +211,8 @@ struct spa { uint8_t spa_sync_on; /* sync threads are running */ spa_load_state_t spa_load_state; /* current load operation */ boolean_t spa_indirect_vdevs_loaded; /* mappings loaded? */ + boolean_t spa_trust_config; /* do we trust vdev tree? */ + spa_config_source_t spa_config_source; /* where config comes from? */ uint64_t spa_import_flags; /* import specific flags */ spa_taskqs_t spa_zio_taskq[ZIO_TYPES][ZIO_TASKQ_TYPES]; dsl_pool_t *spa_dsl_pool; @@ -263,6 +274,8 @@ struct spa { int spa_async_suspended; /* async tasks suspended */ kcondvar_t spa_async_cv; /* wait for thread_exit() */ uint16_t spa_async_tasks; /* async task mask */ + uint64_t spa_missing_tvds; /* unopenable tvds on load */ + uint64_t spa_missing_tvds_allowed; /* allow loading spa? */ spa_removing_phys_t spa_removing_phys; spa_vdev_removal_t *spa_vdev_removal; diff --git a/include/sys/vdev.h b/include/sys/vdev.h index a9b99331b..161e30ae7 100644 --- a/include/sys/vdev.h +++ b/include/sys/vdev.h @@ -48,9 +48,12 @@ typedef enum vdev_dtl_type { extern int zfs_nocacheflush; extern void vdev_dbgmsg(vdev_t *vd, const char *fmt, ...); +extern void vdev_dbgmsg_print_tree(vdev_t *, int); extern int vdev_open(vdev_t *); extern void vdev_open_children(vdev_t *); -extern int vdev_validate(vdev_t *, boolean_t); +extern int vdev_validate(vdev_t *); +extern int vdev_copy_path_strict(vdev_t *, vdev_t *); +extern void vdev_copy_path_relaxed(vdev_t *, vdev_t *); extern void vdev_close(vdev_t *); extern int vdev_create(vdev_t *, uint64_t txg, boolean_t isreplace); extern void vdev_reopen(vdev_t *); @@ -100,6 +103,7 @@ extern void vdev_scan_stat_init(vdev_t *vd); extern void vdev_propagate_state(vdev_t *vd); extern void vdev_set_state(vdev_t *vd, boolean_t isopen, vdev_state_t state, vdev_aux_t aux); +extern boolean_t vdev_children_are_offline(vdev_t *vd); extern void vdev_space_update(vdev_t *vd, int64_t alloc_delta, int64_t defer_delta, int64_t space_delta); @@ -145,7 +149,8 @@ typedef enum vdev_config_flag { VDEV_CONFIG_SPARE = 1 << 0, VDEV_CONFIG_L2CACHE = 1 << 1, VDEV_CONFIG_REMOVING = 1 << 2, - VDEV_CONFIG_MOS = 1 << 3 + VDEV_CONFIG_MOS = 1 << 3, + VDEV_CONFIG_MISSING = 1 << 4 } vdev_config_flag_t; extern void vdev_top_config_generate(spa_t *spa, nvlist_t *config); diff --git a/include/sys/vdev_impl.h b/include/sys/vdev_impl.h index b933f9ab8..e28994613 100644 --- a/include/sys/vdev_impl.h +++ b/include/sys/vdev_impl.h @@ -437,7 +437,6 @@ extern void vdev_remove_parent(vdev_t *cvd); /* * vdev sync load and sync */ -extern void vdev_load_log_state(vdev_t *nvd, vdev_t *ovd); extern boolean_t vdev_log_state_valid(vdev_t *vd); extern int vdev_load(vdev_t *vd); extern int vdev_dtl_load(vdev_t *vd); diff --git a/include/sys/zfs_context.h b/include/sys/zfs_context.h index 2e311cffd..37bdc533c 100644 --- a/include/sys/zfs_context.h +++ b/include/sys/zfs_context.h @@ -674,6 +674,7 @@ typedef struct callb_cpr { #define zone_dataset_visible(x, y) (1) #define INGLOBALZONE(z) (1) +extern uint32_t zone_get_hostid(void *zonep); extern char *kmem_vasprintf(const char *fmt, va_list adx); extern char *kmem_asprintf(const char *fmt, ...); diff --git a/lib/libzfs/libzfs_import.c b/lib/libzfs/libzfs_import.c index cc9a52a3e..68b5988cd 100644 --- a/lib/libzfs/libzfs_import.c +++ b/lib/libzfs/libzfs_import.c @@ -897,7 +897,8 @@ vdev_is_hole(uint64_t *hole_array, uint_t holes, uint_t id) * return to the user. */ static nvlist_t * -get_configs(libzfs_handle_t *hdl, pool_list_t *pl, boolean_t active_ok) +get_configs(libzfs_handle_t *hdl, pool_list_t *pl, boolean_t active_ok, + nvlist_t *policy) { pool_entry_t *pe; vdev_entry_t *ve; @@ -1230,6 +1231,12 @@ get_configs(libzfs_handle_t *hdl, pool_list_t *pl, boolean_t active_ok) continue; } + if (policy != NULL) { + if (nvlist_add_nvlist(config, ZPOOL_REWIND_POLICY, + policy) != 0) + goto nomem; + } + if ((nvl = refresh_config(hdl, config)) == NULL) { nvlist_free(config); config = NULL; @@ -2080,7 +2087,7 @@ zpool_find_import_impl(libzfs_handle_t *hdl, importargs_t *iarg) free(cache); pthread_mutex_destroy(&lock); - ret = get_configs(hdl, &pools, iarg->can_be_active); + ret = get_configs(hdl, &pools, iarg->can_be_active, iarg->policy); for (pe = pools.pools; pe != NULL; pe = penext) { penext = pe->pe_next; @@ -2209,6 +2216,14 @@ zpool_find_import_cached(libzfs_handle_t *hdl, const char *cachefile, if (active) continue; + if (nvlist_add_string(src, ZPOOL_CONFIG_CACHEFILE, + cachefile) != 0) { + (void) no_memory(hdl); + nvlist_free(raw); + nvlist_free(pools); + return (NULL); + } + if ((dst = refresh_config(hdl, src)) == NULL) { nvlist_free(raw); nvlist_free(pools); diff --git a/lib/libzfs/libzfs_pool.c b/lib/libzfs/libzfs_pool.c index 2d94cd320..d082a5f66 100644 --- a/lib/libzfs/libzfs_pool.c +++ b/lib/libzfs/libzfs_pool.c @@ -1935,8 +1935,9 @@ zpool_import_props(libzfs_handle_t *hdl, nvlist_t *config, const char *newname, nvlist_lookup_nvlist(nvinfo, ZPOOL_CONFIG_MISSING_DEVICES, &missing) == 0) { (void) printf(dgettext(TEXT_DOMAIN, - "The devices below are missing, use " - "'-m' to import the pool anyway:\n")); + "The devices below are missing or " + "corrupted, use '-m' to import the pool " + "anyway:\n")); print_vdev_tree(hdl, NULL, missing, 2); (void) printf("\n"); } diff --git a/lib/libzpool/kernel.c b/lib/libzpool/kernel.c index 09e69ef6d..f3e84975c 100644 --- a/lib/libzpool/kernel.c +++ b/lib/libzpool/kernel.c @@ -297,6 +297,16 @@ rw_tryenter(krwlock_t *rwlp, krw_t rw) return (0); } +/* ARGSUSED */ +uint32_t +zone_get_hostid(void *zonep) +{ + /* + * We're emulating the system's hostid in userland. + */ + return (strtoul(hw_serial, NULL, 10)); +} + int rw_tryupgrade(krwlock_t *rwlp) { diff --git a/man/man5/zfs-module-parameters.5 b/man/man5/zfs-module-parameters.5 index 822146a7a..886dffce8 100644 --- a/man/man5/zfs-module-parameters.5 +++ b/man/man5/zfs-module-parameters.5 @@ -354,6 +354,18 @@ Default value: \fB24\fR. .sp .ne 2 .na +\fBspa_load_print_vdev_tree\fR (int) +.ad +.RS 12n +Whether to print the vdev tree in the debugging message buffer during pool import. +Use 0 to disable and 1 to enable. +.sp +Default value: \fB0\fR. +.RE + +.sp +.ne 2 +.na \fBspa_load_verify_data\fR (int) .ad .RS 12n @@ -704,6 +716,18 @@ Default value: \fB0\fR. .sp .ne 2 .na +\fBzfs_max_missing_tvds\fR (int) +.ad +.RS 12n +Number of missing top-level vdevs which will be allowed during +pool import (only in read-only mode). +.sp +Default value: \fB0\fR +.RE + +.sp +.ne 2 +.na \fBzfs_multilist_num_sublists\fR (int) .ad .RS 12n diff --git a/module/zfs/spa.c b/module/zfs/spa.c index 80f0c6f36..3177f9649 100644 --- a/module/zfs/spa.c +++ b/module/zfs/spa.c @@ -157,9 +157,8 @@ const zio_taskq_info_t zio_taskqs[ZIO_TYPES][ZIO_TASKQ_TYPES] = { static void spa_sync_version(void *arg, dmu_tx_t *tx); static void spa_sync_props(void *arg, dmu_tx_t *tx); static boolean_t spa_has_active_shared_spare(spa_t *spa); -static inline int spa_load_impl(spa_t *spa, uint64_t, nvlist_t *config, - spa_load_state_t state, spa_import_type_t type, boolean_t trust_config, - char **ereport); +static int spa_load_impl(spa_t *spa, spa_import_type_t type, char **ereport, + boolean_t reloading); static void spa_vdev_resilver_done(spa_t *spa); uint_t zio_taskq_batch_pct = 75; /* 1 thread per cpu in pset */ @@ -182,6 +181,54 @@ boolean_t spa_load_verify_dryrun = B_FALSE; #define TRYIMPORT_NAME "$import" /* + * For debugging purposes: print out vdev tree during pool import. + */ +int spa_load_print_vdev_tree = B_FALSE; + +/* + * A non-zero value for zfs_max_missing_tvds means that we allow importing + * pools with missing top-level vdevs. This is strictly intended for advanced + * pool recovery cases since missing data is almost inevitable. Pools with + * missing devices can only be imported read-only for safety reasons, and their + * fail-mode will be automatically set to "continue". + * + * With 1 missing vdev we should be able to import the pool and mount all + * datasets. User data that was not modified after the missing device has been + * added should be recoverable. This means that snapshots created prior to the + * addition of that device should be completely intact. + * + * With 2 missing vdevs, some datasets may fail to mount since there are + * dataset statistics that are stored as regular metadata. Some data might be + * recoverable if those vdevs were added recently. + * + * With 3 or more missing vdevs, the pool is severely damaged and MOS entries + * may be missing entirely. Chances of data recovery are very low. Note that + * there are also risks of performing an inadvertent rewind as we might be + * missing all the vdevs with the latest uberblocks. + */ +unsigned long zfs_max_missing_tvds = 0; + +/* + * The parameters below are similar to zfs_max_missing_tvds but are only + * intended for a preliminary open of the pool with an untrusted config which + * might be incomplete or out-dated. + * + * We are more tolerant for pools opened from a cachefile since we could have + * an out-dated cachefile where a device removal was not registered. + * We could have set the limit arbitrarily high but in the case where devices + * are really missing we would want to return the proper error codes; we chose + * SPA_DVAS_PER_BP - 1 so that some copies of the MOS would still be available + * and we get a chance to retrieve the trusted config. + */ +uint64_t zfs_max_missing_tvds_cachefile = SPA_DVAS_PER_BP - 1; +/* + * In the case where config was assembled by scanning device paths (/dev/dsks + * by default) we are less tolerant since all the existing devices should have + * been detected and we want spa_load to return the right error codes. + */ +uint64_t zfs_max_missing_tvds_scan = 0; + +/* * ========================================================================== * SPA properties routines * ========================================================================== @@ -1757,13 +1804,34 @@ load_nvlist(spa_t *spa, uint64_t obj, nvlist_t **value) } /* + * Concrete top-level vdevs that are not missing and are not logs. At every + * spa_sync we write new uberblocks to at least SPA_SYNC_MIN_VDEVS core tvds. + */ +static uint64_t +spa_healthy_core_tvds(spa_t *spa) +{ + vdev_t *rvd = spa->spa_root_vdev; + uint64_t tvds = 0; + + for (uint64_t i = 0; i < rvd->vdev_children; i++) { + vdev_t *vd = rvd->vdev_child[i]; + if (vd->vdev_islog) + continue; + if (vdev_is_concrete(vd) && !vdev_is_dead(vd)) + tvds++; + } + + return (tvds); +} + +/* * Checks to see if the given vdev could not be opened, in which case we post a * sysevent to notify the autoreplace code that the device has been removed. */ static void spa_check_removed(vdev_t *vd) { - for (int c = 0; c < vd->vdev_children; c++) + for (uint64_t c = 0; c < vd->vdev_children; c++) spa_check_removed(vd->vdev_child[c]); if (vd->vdev_ops->vdev_op_leaf && vdev_is_dead(vd) && @@ -1773,38 +1841,14 @@ spa_check_removed(vdev_t *vd) } } -static void -spa_config_valid_zaps(vdev_t *vd, vdev_t *mvd) -{ - ASSERT3U(vd->vdev_children, ==, mvd->vdev_children); - - vd->vdev_top_zap = mvd->vdev_top_zap; - vd->vdev_leaf_zap = mvd->vdev_leaf_zap; - - for (uint64_t i = 0; i < vd->vdev_children; i++) { - spa_config_valid_zaps(vd->vdev_child[i], mvd->vdev_child[i]); - } -} - -/* - * Validate the current config against the MOS config - */ -static boolean_t -spa_config_valid(spa_t *spa, nvlist_t *config) +static int +spa_check_for_missing_logs(spa_t *spa) { - vdev_t *mrvd, *rvd = spa->spa_root_vdev; - nvlist_t *nv; - - VERIFY(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &nv) == 0); - - spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); - VERIFY(spa_config_parse(spa, &mrvd, nv, NULL, 0, VDEV_ALLOC_LOAD) == 0); - - ASSERT3U(rvd->vdev_children, ==, mrvd->vdev_children); + vdev_t *rvd = spa->spa_root_vdev; /* * If we're doing a normal import, then build up any additional - * diagnostic information about missing devices in this config. + * diagnostic information about missing log devices. * We'll pass this up to the user for further processing. */ if (!(spa->spa_import_flags & ZFS_IMPORT_MISSING_LOG)) { @@ -1815,109 +1859,52 @@ spa_config_valid(spa_t *spa, nvlist_t *config) KM_SLEEP); VERIFY(nvlist_alloc(&nv, NV_UNIQUE_NAME, KM_SLEEP) == 0); - for (int c = 0; c < rvd->vdev_children; c++) { + for (uint64_t c = 0; c < rvd->vdev_children; c++) { vdev_t *tvd = rvd->vdev_child[c]; - vdev_t *mtvd = mrvd->vdev_child[c]; - if (tvd->vdev_ops == &vdev_missing_ops && - mtvd->vdev_ops != &vdev_missing_ops && - mtvd->vdev_islog) - child[idx++] = vdev_config_generate(spa, mtvd, - B_FALSE, 0); + /* + * We consider a device as missing only if it failed + * to open (i.e. offline or faulted is not considered + * as missing). + */ + if (tvd->vdev_islog && + tvd->vdev_state == VDEV_STATE_CANT_OPEN) { + child[idx++] = vdev_config_generate(spa, tvd, + B_FALSE, VDEV_CONFIG_MISSING); + } } - if (idx) { - VERIFY(nvlist_add_nvlist_array(nv, - ZPOOL_CONFIG_CHILDREN, child, idx) == 0); - VERIFY(nvlist_add_nvlist(spa->spa_load_info, - ZPOOL_CONFIG_MISSING_DEVICES, nv) == 0); + if (idx > 0) { + fnvlist_add_nvlist_array(nv, + ZPOOL_CONFIG_CHILDREN, child, idx); + fnvlist_add_nvlist(spa->spa_load_info, + ZPOOL_CONFIG_MISSING_DEVICES, nv); - for (int i = 0; i < idx; i++) + for (uint64_t i = 0; i < idx; i++) nvlist_free(child[i]); } nvlist_free(nv); kmem_free(child, rvd->vdev_children * sizeof (char **)); - } - /* - * Compare the root vdev tree with the information we have - * from the MOS config (mrvd). Check each top-level vdev - * with the corresponding MOS config top-level (mtvd). - */ - for (int c = 0; c < rvd->vdev_children; c++) { - vdev_t *tvd = rvd->vdev_child[c]; - vdev_t *mtvd = mrvd->vdev_child[c]; - - /* - * Resolve any "missing" vdevs in the current configuration. - * Also trust the MOS config about any "indirect" vdevs. - * If we find that the MOS config has more accurate information - * about the top-level vdev then use that vdev instead. - */ - if ((tvd->vdev_ops == &vdev_missing_ops && - mtvd->vdev_ops != &vdev_missing_ops) || - (mtvd->vdev_ops == &vdev_indirect_ops && - tvd->vdev_ops != &vdev_indirect_ops)) { - - /* - * Device specific actions. - */ - if (mtvd->vdev_islog) { - if (!(spa->spa_import_flags & - ZFS_IMPORT_MISSING_LOG)) { - continue; - } + if (idx > 0) { + spa_load_failed(spa, "some log devices are missing"); + return (SET_ERROR(ENXIO)); + } + } else { + for (uint64_t c = 0; c < rvd->vdev_children; c++) { + vdev_t *tvd = rvd->vdev_child[c]; + if (tvd->vdev_islog && + tvd->vdev_state == VDEV_STATE_CANT_OPEN) { spa_set_log_state(spa, SPA_LOG_CLEAR); - } else if (mtvd->vdev_ops != &vdev_indirect_ops) { - continue; - } - - /* - * Swap the missing vdev with the data we were - * able to obtain from the MOS config. - */ - vdev_remove_child(rvd, tvd); - vdev_remove_child(mrvd, mtvd); - - vdev_add_child(rvd, mtvd); - vdev_add_child(mrvd, tvd); - - vdev_reopen(rvd); - } else { - if (mtvd->vdev_islog) { - /* - * Load the slog device's state from the MOS - * config since it's possible that the label - * does not contain the most up-to-date - * information. - */ - vdev_load_log_state(tvd, mtvd); - vdev_reopen(tvd); + spa_load_note(spa, "some log devices are " + "missing, ZIL is dropped."); + break; } - - /* - * Per-vdev ZAP info is stored exclusively in the MOS. - */ - spa_config_valid_zaps(tvd, mtvd); } - - /* - * Never trust this info from userland; always use what's - * in the MOS. This prevents it from getting out of sync - * with the rest of the info in the MOS. - */ - tvd->vdev_removing = mtvd->vdev_removing; - tvd->vdev_indirect_config = mtvd->vdev_indirect_config; } - vdev_free(mrvd); - spa_config_exit(spa, SCL_ALL, FTAG); - - /* - * Ensure we were able to validate the config. - */ - return (rvd->vdev_guid_sum == spa->spa_uberblock.ub_guid_sum); + return (0); } /* @@ -2311,53 +2298,15 @@ spa_try_repair(spa_t *spa, nvlist_t *config) } static int -spa_load(spa_t *spa, spa_load_state_t state, spa_import_type_t type, - boolean_t trust_config) +spa_load(spa_t *spa, spa_load_state_t state, spa_import_type_t type) { - nvlist_t *config = spa->spa_config; char *ereport = FM_EREPORT_ZFS_POOL; - char *comment; int error; - uint64_t pool_guid; - nvlist_t *nvl; - - if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, &pool_guid)) - return (SET_ERROR(EINVAL)); - - ASSERT(spa->spa_comment == NULL); - if (nvlist_lookup_string(config, ZPOOL_CONFIG_COMMENT, &comment) == 0) - spa->spa_comment = spa_strdup(comment); - - /* - * Versioning wasn't explicitly added to the label until later, so if - * it's not present treat it as the initial version. - */ - if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, - &spa->spa_ubsync.ub_version) != 0) - spa->spa_ubsync.ub_version = SPA_VERSION_INITIAL; - - (void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_TXG, - &spa->spa_config_txg); - - if ((state == SPA_LOAD_IMPORT || state == SPA_LOAD_TRYIMPORT) && - spa_guid_exists(pool_guid, 0)) { - error = SET_ERROR(EEXIST); - } else { - spa->spa_config_guid = pool_guid; - if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_SPLIT, - &nvl) == 0) { - VERIFY(nvlist_dup(nvl, &spa->spa_config_splitting, - KM_SLEEP) == 0); - } - - nvlist_free(spa->spa_load_info); - spa->spa_load_info = fnvlist_alloc(); + spa->spa_load_state = state; - gethrestime(&spa->spa_loaded_ts); - error = spa_load_impl(spa, pool_guid, config, state, type, - trust_config, &ereport); - } + gethrestime(&spa->spa_loaded_ts); + error = spa_load_impl(spa, type, &ereport, B_FALSE); /* * Don't count references from objsets that are already closed @@ -2611,13 +2560,80 @@ out: } static int -spa_ld_parse_config(spa_t *spa, uint64_t pool_guid, nvlist_t *config, - spa_import_type_t type) +spa_verify_host(spa_t *spa, nvlist_t *mos_config) +{ + uint64_t hostid; + char *hostname; + uint64_t myhostid = 0; + + if (!spa_is_root(spa) && nvlist_lookup_uint64(mos_config, + ZPOOL_CONFIG_HOSTID, &hostid) == 0) { + hostname = fnvlist_lookup_string(mos_config, + ZPOOL_CONFIG_HOSTNAME); + + myhostid = zone_get_hostid(NULL); + + if (hostid != 0 && myhostid != 0 && hostid != myhostid) { + cmn_err(CE_WARN, "pool '%s' could not be " + "loaded as it was last accessed by " + "another system (host: %s hostid: 0x%llx). " + "See: http://illumos.org/msg/ZFS-8000-EY", + spa_name(spa), hostname, (u_longlong_t)hostid); + spa_load_failed(spa, "hostid verification failed: pool " + "last accessed by host: %s (hostid: 0x%llx)", + hostname, (u_longlong_t)hostid); + return (SET_ERROR(EBADF)); + } + } + + return (0); +} + +static int +spa_ld_parse_config(spa_t *spa, spa_import_type_t type) { int error = 0; - nvlist_t *nvtree = NULL; + nvlist_t *nvtree, *nvl, *config = spa->spa_config; int parse; vdev_t *rvd; + uint64_t pool_guid; + char *comment; + + /* + * Versioning wasn't explicitly added to the label until later, so if + * it's not present treat it as the initial version. + */ + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, + &spa->spa_ubsync.ub_version) != 0) + spa->spa_ubsync.ub_version = SPA_VERSION_INITIAL; + + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, &pool_guid)) { + spa_load_failed(spa, "invalid config provided: '%s' missing", + ZPOOL_CONFIG_POOL_GUID); + return (SET_ERROR(EINVAL)); + } + + if ((spa->spa_load_state == SPA_LOAD_IMPORT || spa->spa_load_state == + SPA_LOAD_TRYIMPORT) && spa_guid_exists(pool_guid, 0)) { + spa_load_failed(spa, "a pool with guid %llu is already open", + (u_longlong_t)pool_guid); + return (SET_ERROR(EEXIST)); + } + + spa->spa_config_guid = pool_guid; + + nvlist_free(spa->spa_load_info); + spa->spa_load_info = fnvlist_alloc(); + + ASSERT(spa->spa_comment == NULL); + if (nvlist_lookup_string(config, ZPOOL_CONFIG_COMMENT, &comment) == 0) + spa->spa_comment = spa_strdup(comment); + + (void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_TXG, + &spa->spa_config_txg); + + if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_SPLIT, &nvl) == 0) + spa->spa_config_splitting = fnvlist_dup(nvl); if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &nvtree)) { spa_load_failed(spa, "invalid config provided: '%s' missing", @@ -2625,9 +2641,6 @@ spa_ld_parse_config(spa_t *spa, uint64_t pool_guid, nvlist_t *config, return (SET_ERROR(EINVAL)); } - parse = (type == SPA_IMPORT_EXISTING ? - VDEV_ALLOC_LOAD : VDEV_ALLOC_SPLIT); - /* * Create "The Godfather" zio to hold all async IOs */ @@ -2645,6 +2658,8 @@ spa_ld_parse_config(spa_t *spa, uint64_t pool_guid, nvlist_t *config, * configuration requires knowing the version number. */ spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); + parse = (type == SPA_IMPORT_EXISTING ? + VDEV_ALLOC_LOAD : VDEV_ALLOC_SPLIT); error = spa_config_parse(spa, &rvd, nvtree, NULL, 0, parse); spa_config_exit(spa, SCL_ALL, FTAG); @@ -2665,71 +2680,105 @@ spa_ld_parse_config(spa_t *spa, uint64_t pool_guid, nvlist_t *config, return (0); } +/* + * Recursively open all vdevs in the vdev tree. This function is called twice: + * first with the untrusted config, then with the trusted config. + */ static int spa_ld_open_vdevs(spa_t *spa) { int error = 0; + /* + * spa_missing_tvds_allowed defines how many top-level vdevs can be + * missing/unopenable for the root vdev to be still considered openable. + */ + if (spa->spa_trust_config) { + spa->spa_missing_tvds_allowed = zfs_max_missing_tvds; + } else if (spa->spa_config_source == SPA_CONFIG_SRC_CACHEFILE) { + spa->spa_missing_tvds_allowed = zfs_max_missing_tvds_cachefile; + } else if (spa->spa_config_source == SPA_CONFIG_SRC_SCAN) { + spa->spa_missing_tvds_allowed = zfs_max_missing_tvds_scan; + } else { + spa->spa_missing_tvds_allowed = 0; + } + + spa->spa_missing_tvds_allowed = + MAX(zfs_max_missing_tvds, spa->spa_missing_tvds_allowed); + spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); error = vdev_open(spa->spa_root_vdev); spa_config_exit(spa, SCL_ALL, FTAG); + + if (spa->spa_missing_tvds != 0) { + spa_load_note(spa, "vdev tree has %lld missing top-level " + "vdevs.", (u_longlong_t)spa->spa_missing_tvds); + if (spa->spa_trust_config && (spa->spa_mode & FWRITE)) { + /* + * Although theoretically we could allow users to open + * incomplete pools in RW mode, we'd need to add a lot + * of extra logic (e.g. adjust pool space to account + * for missing vdevs). + * This limitation also prevents users from accidentally + * opening the pool in RW mode during data recovery and + * damaging it further. + */ + spa_load_note(spa, "pools with missing top-level " + "vdevs can only be opened in read-only mode."); + error = SET_ERROR(ENXIO); + } else { + spa_load_note(spa, "current settings allow for maximum " + "%lld missing top-level vdevs at this stage.", + (u_longlong_t)spa->spa_missing_tvds_allowed); + } + } if (error != 0) { spa_load_failed(spa, "unable to open vdev tree [error=%d]", error); } + if (spa->spa_missing_tvds != 0 || error != 0) + vdev_dbgmsg_print_tree(spa->spa_root_vdev, 2); return (error); } +/* + * We need to validate the vdev labels against the configuration that + * we have in hand. This function is called twice: first with an untrusted + * config, then with a trusted config. The validation is more strict when the + * config is trusted. + */ static int -spa_ld_validate_vdevs(spa_t *spa, spa_import_type_t type, - boolean_t trust_config) +spa_ld_validate_vdevs(spa_t *spa) { int error = 0; vdev_t *rvd = spa->spa_root_vdev; - /* - * We need to validate the vdev labels against the configuration that - * we have in hand, which is dependent on the setting of trust_config. - * If trust_config is true then we're validating the vdev labels based - * on that config. Otherwise, we're validating against the cached - * config (zpool.cache) that was read when we loaded the zfs module, and - * then later we will recursively call spa_load() and validate against - * the vdev config. - * - * If we're assembling a new pool that's been split off from an - * existing pool, the labels haven't yet been updated so we skip - * validation for now. - */ - if (type != SPA_IMPORT_ASSEMBLE) { - spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); - error = vdev_validate(rvd, trust_config); - spa_config_exit(spa, SCL_ALL, FTAG); + spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); + error = vdev_validate(rvd); + spa_config_exit(spa, SCL_ALL, FTAG); - if (error != 0) { - spa_load_failed(spa, "vdev_validate failed [error=%d]", - error); - return (error); - } + if (error != 0) { + spa_load_failed(spa, "vdev_validate failed [error=%d]", error); + return (error); + } - if (rvd->vdev_state <= VDEV_STATE_CANT_OPEN) { - spa_load_failed(spa, "cannot open vdev tree after " - "invalidating some vdevs"); - return (SET_ERROR(ENXIO)); - } + if (rvd->vdev_state <= VDEV_STATE_CANT_OPEN) { + spa_load_failed(spa, "cannot open vdev tree after invalidating " + "some vdevs"); + vdev_dbgmsg_print_tree(rvd, 2); + return (SET_ERROR(ENXIO)); } return (0); } static int -spa_ld_select_uberblock(spa_t *spa, nvlist_t *config, spa_import_type_t type, - boolean_t trust_config) +spa_ld_select_uberblock(spa_t *spa, spa_import_type_t type) { vdev_t *rvd = spa->spa_root_vdev; nvlist_t *label; uberblock_t *ub = &spa->spa_uberblock; - uint64_t children; boolean_t activity_check = B_FALSE; /* @@ -2755,7 +2804,8 @@ spa_ld_select_uberblock(spa_t *spa, nvlist_t *config, spa_import_type_t type, * pool is truly inactive and can be safely imported. Prevent * hosts which don't have a hostid set from importing the pool. */ - activity_check = spa_activity_check_required(spa, ub, label, config); + activity_check = spa_activity_check_required(spa, ub, label, + spa->spa_config); if (activity_check) { if (ub->ub_mmp_magic == MMP_MAGIC && ub->ub_mmp_delay && spa_get_hostid() == 0) { @@ -2765,7 +2815,7 @@ spa_ld_select_uberblock(spa_t *spa, nvlist_t *config, spa_import_type_t type, return (spa_vdev_err(rvd, VDEV_AUX_ACTIVE, EREMOTEIO)); } - int error = spa_activity_check(spa, ub, config); + int error = spa_activity_check(spa, ub, spa->spa_config); if (error) { nvlist_free(label); return (error); @@ -2851,26 +2901,9 @@ spa_ld_select_uberblock(spa_t *spa, nvlist_t *config, spa_import_type_t type, nvlist_free(unsup_feat); } - /* - * If the vdev guid sum doesn't match the uberblock, we have an - * incomplete configuration. We first check to see if the pool - * is aware of the complete config (i.e ZPOOL_CONFIG_VDEV_CHILDREN). - * If it is, defer the vdev_guid_sum check till later so we - * can handle missing vdevs. - */ - if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_VDEV_CHILDREN, - &children) != 0 && trust_config && type != SPA_IMPORT_ASSEMBLE && - rvd->vdev_guid_sum != ub->ub_guid_sum) { - spa_load_failed(spa, "guid sum in config doesn't match guid " - "sum in uberblock (%llu != %llu)", - (u_longlong_t)rvd->vdev_guid_sum, - (u_longlong_t)ub->ub_guid_sum); - return (spa_vdev_err(rvd, VDEV_AUX_BAD_GUID_SUM, ENXIO)); - } - if (type != SPA_IMPORT_ASSEMBLE && spa->spa_config_splitting) { spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); - spa_try_repair(spa, config); + spa_try_repair(spa, spa->spa_config); spa_config_exit(spa, SCL_ALL, FTAG); nvlist_free(spa->spa_config_splitting); spa->spa_config_splitting = NULL; @@ -2909,49 +2942,167 @@ spa_ld_open_rootbp(spa_t *spa) } static int -spa_ld_validate_config(spa_t *spa, spa_import_type_t type) +spa_ld_load_trusted_config(spa_t *spa, spa_import_type_t type, + boolean_t reloading) { - vdev_t *rvd = spa->spa_root_vdev; + vdev_t *mrvd, *rvd = spa->spa_root_vdev; + nvlist_t *nv, *mos_config, *policy; + int error = 0, copy_error; + uint64_t healthy_tvds, healthy_tvds_mos; + uint64_t mos_config_txg; if (spa_dir_prop(spa, DMU_POOL_CONFIG, &spa->spa_config_object, B_TRUE) != 0) return (spa_vdev_err(rvd, VDEV_AUX_CORRUPT_DATA, EIO)); /* - * Validate the config, using the MOS config to fill in any - * information which might be missing. If we fail to validate - * the config then declare the pool unfit for use. If we're - * assembling a pool from a split, the log is not transferred - * over. + * If we're assembling a pool from a split, the config provided is + * already trusted so there is nothing to do. */ - if (type != SPA_IMPORT_ASSEMBLE) { - nvlist_t *mos_config; - if (load_nvlist(spa, spa->spa_config_object, &mos_config) - != 0) { - spa_load_failed(spa, "unable to retrieve MOS config"); - return (spa_vdev_err(rvd, VDEV_AUX_CORRUPT_DATA, EIO)); - } + if (type == SPA_IMPORT_ASSEMBLE) + return (0); + + healthy_tvds = spa_healthy_core_tvds(spa); - if (!spa_config_valid(spa, mos_config)) { + if (load_nvlist(spa, spa->spa_config_object, &mos_config) + != 0) { + spa_load_failed(spa, "unable to retrieve MOS config"); + return (spa_vdev_err(rvd, VDEV_AUX_CORRUPT_DATA, EIO)); + } + + /* + * If we are doing an open, pool owner wasn't verified yet, thus do + * the verification here. + */ + if (spa->spa_load_state == SPA_LOAD_OPEN) { + error = spa_verify_host(spa, mos_config); + if (error != 0) { nvlist_free(mos_config); - spa_load_failed(spa, "mismatch between config provided " - "and config stored in MOS"); - return (spa_vdev_err(rvd, VDEV_AUX_BAD_GUID_SUM, - ENXIO)); + return (error); } - nvlist_free(mos_config); + } + + nv = fnvlist_lookup_nvlist(mos_config, ZPOOL_CONFIG_VDEV_TREE); + spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); + + /* + * Build a new vdev tree from the trusted config + */ + VERIFY(spa_config_parse(spa, &mrvd, nv, NULL, 0, VDEV_ALLOC_LOAD) == 0); + + /* + * Vdev paths in the MOS may be obsolete. If the untrusted config was + * obtained by scanning /dev/dsk, then it will have the right vdev + * paths. We update the trusted MOS config with this information. + * We first try to copy the paths with vdev_copy_path_strict, which + * succeeds only when both configs have exactly the same vdev tree. + * If that fails, we fall back to a more flexible method that has a + * best effort policy. + */ + copy_error = vdev_copy_path_strict(rvd, mrvd); + if (copy_error != 0 || spa_load_print_vdev_tree) { + spa_load_note(spa, "provided vdev tree:"); + vdev_dbgmsg_print_tree(rvd, 2); + spa_load_note(spa, "MOS vdev tree:"); + vdev_dbgmsg_print_tree(mrvd, 2); + } + if (copy_error != 0) { + spa_load_note(spa, "vdev_copy_path_strict failed, falling " + "back to vdev_copy_path_relaxed"); + vdev_copy_path_relaxed(rvd, mrvd); + } + + vdev_close(rvd); + vdev_free(rvd); + spa->spa_root_vdev = mrvd; + rvd = mrvd; + spa_config_exit(spa, SCL_ALL, FTAG); + + /* + * We will use spa_config if we decide to reload the spa or if spa_load + * fails and we rewind. We must thus regenerate the config using the + * MOS information with the updated paths. Rewind policy is an import + * setting and is not in the MOS. We copy it over to our new, trusted + * config. + */ + mos_config_txg = fnvlist_lookup_uint64(mos_config, + ZPOOL_CONFIG_POOL_TXG); + nvlist_free(mos_config); + mos_config = spa_config_generate(spa, NULL, mos_config_txg, B_FALSE); + if (nvlist_lookup_nvlist(spa->spa_config, ZPOOL_REWIND_POLICY, + &policy) == 0) + fnvlist_add_nvlist(mos_config, ZPOOL_REWIND_POLICY, policy); + spa_config_set(spa, mos_config); + spa->spa_config_source = SPA_CONFIG_SRC_MOS; + + /* + * Now that we got the config from the MOS, we should be more strict + * in checking blkptrs and can make assumptions about the consistency + * of the vdev tree. spa_trust_config must be set to true before opening + * vdevs in order for them to be writeable. + */ + spa->spa_trust_config = B_TRUE; + + /* + * Open and validate the new vdev tree + */ + error = spa_ld_open_vdevs(spa); + if (error != 0) + return (error); + + error = spa_ld_validate_vdevs(spa); + if (error != 0) + return (error); + + if (copy_error != 0 || spa_load_print_vdev_tree) { + spa_load_note(spa, "final vdev tree:"); + vdev_dbgmsg_print_tree(rvd, 2); + } + + if (spa->spa_load_state != SPA_LOAD_TRYIMPORT && + !spa->spa_extreme_rewind && zfs_max_missing_tvds == 0) { /* - * Now that we've validated the config, check the state of the - * root vdev. If it can't be opened, it indicates one or - * more toplevel vdevs are faulted. + * Sanity check to make sure that we are indeed loading the + * latest uberblock. If we missed SPA_SYNC_MIN_VDEVS tvds + * in the config provided and they happened to be the only ones + * to have the latest uberblock, we could involuntarily perform + * an extreme rewind. */ - if (rvd->vdev_state <= VDEV_STATE_CANT_OPEN) { - spa_load_failed(spa, "some top vdevs are unavailable"); - return (SET_ERROR(ENXIO)); + healthy_tvds_mos = spa_healthy_core_tvds(spa); + if (healthy_tvds_mos - healthy_tvds >= + SPA_SYNC_MIN_VDEVS) { + spa_load_note(spa, "config provided misses too many " + "top-level vdevs compared to MOS (%lld vs %lld). ", + (u_longlong_t)healthy_tvds, + (u_longlong_t)healthy_tvds_mos); + spa_load_note(spa, "vdev tree:"); + vdev_dbgmsg_print_tree(rvd, 2); + if (reloading) { + spa_load_failed(spa, "config was already " + "provided from MOS. Aborting."); + return (spa_vdev_err(rvd, + VDEV_AUX_CORRUPT_DATA, EIO)); + } + spa_load_note(spa, "spa must be reloaded using MOS " + "config"); + return (SET_ERROR(EAGAIN)); } } + error = spa_check_for_missing_logs(spa); + if (error != 0) + return (spa_vdev_err(rvd, VDEV_AUX_BAD_GUID_SUM, ENXIO)); + + if (rvd->vdev_guid_sum != spa->spa_uberblock.ub_guid_sum) { + spa_load_failed(spa, "uberblock guid sum doesn't match MOS " + "guid sum (%llu != %llu)", + (u_longlong_t)spa->spa_uberblock.ub_guid_sum, + (u_longlong_t)rvd->vdev_guid_sum); + return (spa_vdev_err(rvd, VDEV_AUX_BAD_GUID_SUM, + ENXIO)); + } + return (0); } @@ -3118,47 +3269,6 @@ spa_ld_load_special_directories(spa_t *spa) } static int -spa_ld_prepare_for_reload(spa_t *spa, int orig_mode) -{ - vdev_t *rvd = spa->spa_root_vdev; - - uint64_t hostid; - nvlist_t *policy = NULL; - nvlist_t *mos_config; - - if (load_nvlist(spa, spa->spa_config_object, &mos_config) != 0) { - spa_load_failed(spa, "unable to retrieve MOS config"); - return (spa_vdev_err(rvd, VDEV_AUX_CORRUPT_DATA, EIO)); - } - - if (!spa_is_root(spa) && nvlist_lookup_uint64(mos_config, - ZPOOL_CONFIG_HOSTID, &hostid) == 0) { - char *hostname; - unsigned long myhostid = 0; - - VERIFY(nvlist_lookup_string(mos_config, - ZPOOL_CONFIG_HOSTNAME, &hostname) == 0); - - myhostid = spa_get_hostid(); - if (hostid && myhostid && hostid != myhostid) { - nvlist_free(mos_config); - return (SET_ERROR(EBADF)); - } - } - if (nvlist_lookup_nvlist(spa->spa_config, - ZPOOL_REWIND_POLICY, &policy) == 0) - VERIFY(nvlist_add_nvlist(mos_config, - ZPOOL_REWIND_POLICY, policy) == 0); - - spa_config_set(spa, mos_config); - spa_unload(spa); - spa_deactivate(spa); - spa_activate(spa, orig_mode); - - return (0); -} - -static int spa_ld_get_props(spa_t *spa) { int error = 0; @@ -3286,6 +3396,19 @@ spa_ld_get_props(spa_t *spa) spa->spa_autoreplace = (autoreplace != 0); } + /* + * If we are importing a pool with missing top-level vdevs, + * we enforce that the pool doesn't panic or get suspended on + * error since the likelihood of missing data is extremely high. + */ + if (spa->spa_missing_tvds > 0 && + spa->spa_failmode != ZIO_FAILURE_MODE_CONTINUE && + spa->spa_load_state != SPA_LOAD_TRYIMPORT) { + spa_load_note(spa, "forcing failmode to 'continue' " + "as some top level vdevs are missing"); + spa->spa_failmode = ZIO_FAILURE_MODE_CONTINUE; + } + return (0); } @@ -3428,9 +3551,15 @@ spa_ld_verify_logs(spa_t *spa, spa_import_type_t type, char **ereport) if (type != SPA_IMPORT_ASSEMBLE && spa_writeable(spa)) { boolean_t missing = spa_check_logs(spa); if (missing) { - *ereport = FM_EREPORT_ZFS_LOG_REPLAY; - spa_load_failed(spa, "spa_check_logs failed"); - return (spa_vdev_err(rvd, VDEV_AUX_BAD_LOG, ENXIO)); + if (spa->spa_missing_tvds != 0) { + spa_load_note(spa, "spa_check_logs failed " + "so dropping the logs"); + } else { + *ereport = FM_EREPORT_ZFS_LOG_REPLAY; + spa_load_failed(spa, "spa_check_logs failed"); + return (spa_vdev_err(rvd, VDEV_AUX_BAD_LOG, + ENXIO)); + } } } @@ -3486,7 +3615,8 @@ spa_ld_claim_log_blocks(spa_t *spa) } static void -spa_ld_check_for_config_update(spa_t *spa, uint64_t config_cache_txg) +spa_ld_check_for_config_update(spa_t *spa, uint64_t config_cache_txg, + boolean_t reloading) { vdev_t *rvd = spa->spa_root_vdev; int need_update = B_FALSE; @@ -3498,7 +3628,7 @@ spa_ld_check_for_config_update(spa_t *spa, uint64_t config_cache_txg) * If this is a verbatim import, trust the current * in-core spa_config and update the disk labels. */ - if (config_cache_txg != spa->spa_config_txg || + if (reloading || config_cache_txg != spa->spa_config_txg || spa->spa_load_state == SPA_LOAD_IMPORT || spa->spa_load_state == SPA_LOAD_RECOVER || (spa->spa_import_flags & ZFS_IMPORT_VERBATIM)) @@ -3516,6 +3646,24 @@ spa_ld_check_for_config_update(spa_t *spa, uint64_t config_cache_txg) spa_async_request(spa, SPA_ASYNC_CONFIG_UPDATE); } +static void +spa_ld_prepare_for_reload(spa_t *spa) +{ + int mode = spa->spa_mode; + int async_suspended = spa->spa_async_suspended; + + spa_unload(spa); + spa_deactivate(spa); + spa_activate(spa, mode); + + /* + * We save the value of spa_async_suspended as it gets reset to 0 by + * spa_unload(). We want to restore it back to the original value before + * returning as we might be calling spa_async_resume() later. + */ + spa->spa_async_suspended = async_suspended; +} + /* * Load an existing storage pool, using the config provided. This config * describes which vdevs are part of the pool and is later validated against @@ -3523,32 +3671,35 @@ spa_ld_check_for_config_update(spa_t *spa, uint64_t config_cache_txg) * config stored in the MOS. */ static int -spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, - spa_load_state_t state, spa_import_type_t type, boolean_t trust_config, - char **ereport) +spa_load_impl(spa_t *spa, spa_import_type_t type, char **ereport, + boolean_t reloading) { int error = 0; - uint64_t config_cache_txg = spa->spa_config_txg; - int orig_mode = spa->spa_mode; boolean_t missing_feat_write = B_FALSE; ASSERT(MUTEX_HELD(&spa_namespace_lock)); - - spa->spa_load_state = state; - spa_load_note(spa, "LOADING"); + ASSERT(spa->spa_config_source != SPA_CONFIG_SRC_NONE); /* - * If this is an untrusted config, first access the pool in read-only - * mode. We will then retrieve a trusted copy of the config from the MOS - * and use it to reopen the pool in read-write mode. + * Never trust the config that is provided unless we are assembling + * a pool following a split. + * This means don't trust blkptrs and the vdev tree in general. This + * also effectively puts the spa in read-only mode since + * spa_writeable() checks for spa_trust_config to be true. + * We will later load a trusted config from the MOS. */ - if (!trust_config) - spa->spa_mode = FREAD; + if (type != SPA_IMPORT_ASSEMBLE) + spa->spa_trust_config = B_FALSE; + + if (reloading) + spa_load_note(spa, "RELOADING"); + else + spa_load_note(spa, "LOADING"); /* * Parse the config provided to create a vdev tree. */ - error = spa_ld_parse_config(spa, pool_guid, config, type); + error = spa_ld_parse_config(spa, type); if (error != 0) return (error); @@ -3566,10 +3717,15 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, /* * Read the label of each vdev and make sure that the GUIDs stored * there match the GUIDs in the config provided. + * If we're assembling a new pool that's been split off from an + * existing pool, the labels haven't yet been updated so we skip + * validation for now. */ - error = spa_ld_validate_vdevs(spa, type, trust_config); - if (error != 0) - return (error); + if (type != SPA_IMPORT_ASSEMBLE) { + error = spa_ld_validate_vdevs(spa); + if (error != 0) + return (error); + } /* * Read vdev labels to find the best uberblock (i.e. latest, unless @@ -3578,7 +3734,7 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, * label with the best uberblock and verify that our version of zfs * supports them all. */ - error = spa_ld_select_uberblock(spa, config, type, trust_config); + error = spa_ld_select_uberblock(spa, type); if (error != 0) return (error); @@ -3592,13 +3748,21 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, return (error); /* - * Retrieve the config stored in the MOS and use it to validate the - * config provided. Also extract some information from the MOS config - * to update our vdev tree. + * Retrieve the trusted config stored in the MOS and use it to create + * a new, exact version of the vdev tree, then reopen all vdevs. */ - error = spa_ld_validate_config(spa, type); - if (error != 0) + error = spa_ld_load_trusted_config(spa, type, reloading); + if (error == EAGAIN) { + VERIFY(!reloading); + /* + * Redo the loading process with the trusted config if it is + * too different from the untrusted config. + */ + spa_ld_prepare_for_reload(spa); + return (spa_load_impl(spa, type, ereport, B_TRUE)); + } else if (error != 0) { return (error); + } /* * Retrieve the mapping of indirect vdevs. Those vdevs were removed @@ -3629,19 +3793,6 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, return (error); /* - * If the config provided is not trusted, discard it and use the config - * from the MOS to reload the pool. - */ - if (!trust_config) { - error = spa_ld_prepare_for_reload(spa, orig_mode); - if (error != 0) - return (error); - - spa_load_note(spa, "RELOADING"); - return (spa_load(spa, state, SPA_IMPORT_EXISTING, B_TRUE)); - } - - /* * Retrieve pool properties from the MOS. */ error = spa_ld_get_props(spa); @@ -3677,7 +3828,7 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, return (error); if (missing_feat_write) { - ASSERT(state == SPA_LOAD_TRYIMPORT); + ASSERT(spa->spa_load_state == SPA_LOAD_TRYIMPORT); /* * At this point, we know that we can open the pool in @@ -3709,9 +3860,11 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, * pool. If we are importing the pool in read-write mode, a few * additional steps must be performed to finish the import. */ - if (spa_writeable(spa) && (state == SPA_LOAD_RECOVER || + if (spa_writeable(spa) && (spa->spa_load_state == SPA_LOAD_RECOVER || spa->spa_load_max_txg == UINT64_MAX)) { - ASSERT(state != SPA_LOAD_TRYIMPORT); + uint64_t config_cache_txg = spa->spa_config_txg; + + ASSERT(spa->spa_load_state != SPA_LOAD_TRYIMPORT); /* * Traverse the ZIL and claim all blocks. @@ -3739,7 +3892,8 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, * next sync, we would update the config stored in vdev labels * and the cachefile (by default /etc/zfs/zpool.cache). */ - spa_ld_check_for_config_update(spa, config_cache_txg); + spa_ld_check_for_config_update(spa, config_cache_txg, + reloading); /* * Check all DTLs to see if anything needs resilvering. @@ -3776,7 +3930,7 @@ spa_load_impl(spa_t *spa, uint64_t pool_guid, nvlist_t *config, } static int -spa_load_retry(spa_t *spa, spa_load_state_t state, int trust_config) +spa_load_retry(spa_t *spa, spa_load_state_t state) { int mode = spa->spa_mode; @@ -3791,7 +3945,7 @@ spa_load_retry(spa_t *spa, spa_load_state_t state, int trust_config) spa_load_note(spa, "spa_load_retry: rewind, max txg: %llu", (u_longlong_t)spa->spa_load_max_txg); - return (spa_load(spa, state, SPA_IMPORT_EXISTING, trust_config)); + return (spa_load(spa, state, SPA_IMPORT_EXISTING)); } /* @@ -3802,8 +3956,8 @@ spa_load_retry(spa_t *spa, spa_load_state_t state, int trust_config) * spa_load(). */ static int -spa_load_best(spa_t *spa, spa_load_state_t state, int trust_config, - uint64_t max_request, int rewind_flags) +spa_load_best(spa_t *spa, spa_load_state_t state, uint64_t max_request, + int rewind_flags) { nvlist_t *loadinfo = NULL; nvlist_t *config = NULL; @@ -3820,8 +3974,7 @@ spa_load_best(spa_t *spa, spa_load_state_t state, int trust_config, spa->spa_extreme_rewind = B_TRUE; } - load_error = rewind_error = spa_load(spa, state, SPA_IMPORT_EXISTING, - trust_config); + load_error = rewind_error = spa_load(spa, state, SPA_IMPORT_EXISTING); if (load_error == 0) return (0); @@ -3862,7 +4015,7 @@ spa_load_best(spa_t *spa, spa_load_state_t state, int trust_config, spa->spa_uberblock.ub_txg <= spa->spa_load_max_txg) { if (spa->spa_load_max_txg < safe_rewind_txg) spa->spa_extreme_rewind = B_TRUE; - rewind_error = spa_load_retry(spa, state, trust_config); + rewind_error = spa_load_retry(spa, state); } spa->spa_extreme_rewind = B_FALSE; @@ -3944,9 +4097,10 @@ spa_open_common(const char *pool, spa_t **spapp, void *tag, nvlist_t *nvpolicy, if (state != SPA_LOAD_RECOVER) spa->spa_last_ubsync_txg = spa->spa_load_txg = 0; + spa->spa_config_source = SPA_CONFIG_SRC_CACHEFILE; zfs_dbgmsg("spa_open_common: opening %s", pool); - error = spa_load_best(spa, state, B_FALSE, policy.zrp_txg, + error = spa_load_best(spa, state, policy.zrp_txg, policy.zrp_request); if (error == EBADF) { @@ -4863,18 +5017,16 @@ spa_import(char *pool, nvlist_t *config, nvlist_t *props, uint64_t flags) if (policy.zrp_request & ZPOOL_DO_REWIND) state = SPA_LOAD_RECOVER; - /* - * Pass off the heavy lifting to spa_load(). Pass TRUE for trust_config - * because the user-supplied config is actually the one to trust when - * doing an import. - */ - if (state != SPA_LOAD_RECOVER) - spa->spa_last_ubsync_txg = spa->spa_load_txg = 0; + spa->spa_config_source = SPA_CONFIG_SRC_TRYIMPORT; - zfs_dbgmsg("spa_import: importing %s%s", pool, - (state == SPA_LOAD_RECOVER) ? " (RECOVERY MODE)" : ""); - error = spa_load_best(spa, state, B_TRUE, policy.zrp_txg, - policy.zrp_request); + if (state != SPA_LOAD_RECOVER) { + spa->spa_last_ubsync_txg = spa->spa_load_txg = 0; + zfs_dbgmsg("spa_import: importing %s", pool); + } else { + zfs_dbgmsg("spa_import: importing %s, max_txg=%lld " + "(RECOVERY MODE)", pool, (longlong_t)policy.zrp_txg); + } + error = spa_load_best(spa, state, policy.zrp_txg, policy.zrp_request); /* * Propagate anything learned while loading the pool and pass it @@ -4988,10 +5140,11 @@ nvlist_t * spa_tryimport(nvlist_t *tryconfig) { nvlist_t *config = NULL; - char *poolname; + char *poolname, *cachefile; spa_t *spa; uint64_t state; int error; + zpool_rewind_policy_t policy; if (nvlist_lookup_string(tryconfig, ZPOOL_CONFIG_POOL_NAME, &poolname)) return (NULL); @@ -5006,14 +5159,30 @@ spa_tryimport(nvlist_t *tryconfig) spa = spa_add(TRYIMPORT_NAME, tryconfig, NULL); spa_activate(spa, FREAD); - zfs_dbgmsg("spa_tryimport: importing %s", poolname); - /* - * Pass off the heavy lifting to spa_load(). - * Pass TRUE for trust_config because the user-supplied config - * is actually the one to trust when doing an import. + * Rewind pool if a max txg was provided. Note that even though we + * retrieve the complete rewind policy, only the rewind txg is relevant + * for tryimport. */ - error = spa_load(spa, SPA_LOAD_TRYIMPORT, SPA_IMPORT_EXISTING, B_TRUE); + zpool_get_rewind_policy(spa->spa_config, &policy); + if (policy.zrp_txg != UINT64_MAX) { + spa->spa_load_max_txg = policy.zrp_txg; + spa->spa_extreme_rewind = B_TRUE; + zfs_dbgmsg("spa_tryimport: importing %s, max_txg=%lld", + poolname, (longlong_t)policy.zrp_txg); + } else { + zfs_dbgmsg("spa_tryimport: importing %s", poolname); + } + + if (nvlist_lookup_string(tryconfig, ZPOOL_CONFIG_CACHEFILE, &cachefile) + == 0) { + zfs_dbgmsg("spa_tryimport: using cachefile '%s'", cachefile); + spa->spa_config_source = SPA_CONFIG_SRC_CACHEFILE; + } else { + spa->spa_config_source = SPA_CONFIG_SRC_SCAN; + } + + error = spa_load(spa, SPA_LOAD_TRYIMPORT, SPA_IMPORT_EXISTING); /* * If 'tryconfig' was at least parsable, return the current config. @@ -6033,8 +6202,10 @@ spa_vdev_split_mirror(spa_t *spa, char *newname, nvlist_t *config, spa_activate(newspa, spa_mode_global); spa_async_suspend(newspa); + newspa->spa_config_source = SPA_CONFIG_SRC_SPLIT; + /* create the new pool from the disks of the original pool */ - error = spa_load(newspa, SPA_LOAD_IMPORT, SPA_IMPORT_ASSEMBLE, B_TRUE); + error = spa_load(newspa, SPA_LOAD_IMPORT, SPA_IMPORT_ASSEMBLE); if (error) goto out; @@ -7337,7 +7508,7 @@ spa_sync(spa_t *spa, uint64_t txg) spa_config_enter(spa, SCL_STATE, FTAG, RW_READER); if (list_is_empty(&spa->spa_config_dirty_list)) { - vdev_t *svd[SPA_DVAS_PER_BP]; + vdev_t *svd[SPA_SYNC_MIN_VDEVS]; int svdcount = 0; int children = rvd->vdev_children; int c0 = spa_get_random(children); @@ -7348,7 +7519,7 @@ spa_sync(spa_t *spa, uint64_t txg) !vdev_is_concrete(vd)) continue; svd[svdcount++] = vd; - if (svdcount == SPA_DVAS_PER_BP) + if (svdcount == SPA_SYNC_MIN_VDEVS) break; } error = vdev_config_sync(svd, svdcount, txg); @@ -7692,9 +7863,20 @@ module_param(spa_load_verify_data, int, 0644); MODULE_PARM_DESC(spa_load_verify_data, "Set to traverse data on pool import"); +module_param(spa_load_print_vdev_tree, int, 0644); +MODULE_PARM_DESC(spa_load_print_vdev_tree, + "Print vdev tree to zfs_dbgmsg during pool import"); + /* CSTYLED */ module_param(zio_taskq_batch_pct, uint, 0444); MODULE_PARM_DESC(zio_taskq_batch_pct, "Percentage of CPUs to run an IO worker thread"); +/* BEGIN CSTYLED */ +module_param(zfs_max_missing_tvds, ulong, 0644); +MODULE_PARM_DESC(zfs_max_missing_tvds, + "Allow importing pool with up to this number of missing top-level vdevs" + " (in read-only mode)"); +/* END CSTYLED */ + #endif diff --git a/module/zfs/spa_config.c b/module/zfs/spa_config.c index 4e9fd6c57..50bba2345 100644 --- a/module/zfs/spa_config.c +++ b/module/zfs/spa_config.c @@ -393,7 +393,8 @@ void spa_config_set(spa_t *spa, nvlist_t *config) { mutex_enter(&spa->spa_props_lock); - nvlist_free(spa->spa_config); + if (spa->spa_config != NULL && spa->spa_config != config) + nvlist_free(spa->spa_config); spa->spa_config = config; mutex_exit(&spa->spa_props_lock); } diff --git a/module/zfs/spa_misc.c b/module/zfs/spa_misc.c index 6f4db76c8..e0edba155 100644 --- a/module/zfs/spa_misc.c +++ b/module/zfs/spa_misc.c @@ -384,7 +384,8 @@ spa_load_failed(spa_t *spa, const char *fmt, ...) (void) vsnprintf(buf, sizeof (buf), fmt, adx); va_end(adx); - zfs_dbgmsg("spa_load(%s): FAILED: %s", spa->spa_name, buf); + zfs_dbgmsg("spa_load(%s, config %s): FAILED: %s", spa->spa_name, + spa->spa_trust_config ? "trusted" : "untrusted", buf); } /*PRINTFLIKE2*/ @@ -398,7 +399,8 @@ spa_load_note(spa_t *spa, const char *fmt, ...) (void) vsnprintf(buf, sizeof (buf), fmt, adx); va_end(adx); - zfs_dbgmsg("spa_load(%s): %s", spa->spa_name, buf); + zfs_dbgmsg("spa_load(%s, config %s): %s", spa->spa_name, + spa->spa_trust_config ? "trusted" : "untrusted", buf); } /* @@ -637,6 +639,7 @@ spa_add(const char *name, nvlist_t *config, const char *altroot) spa->spa_load_max_txg = UINT64_MAX; spa->spa_proc = &p0; spa->spa_proc_state = SPA_PROC_NONE; + spa->spa_trust_config = B_TRUE; spa->spa_deadman_synctime = MSEC2NSEC(zfs_deadman_synctime_ms); spa->spa_deadman_ziotime = MSEC2NSEC(zfs_deadman_ziotime_ms); @@ -2052,7 +2055,7 @@ spa_is_root(spa_t *spa) boolean_t spa_writeable(spa_t *spa) { - return (!!(spa->spa_mode & FWRITE)); + return (!!(spa->spa_mode & FWRITE) && spa->spa_trust_config); } /* @@ -2233,6 +2236,24 @@ spa_get_hostid(void) return (myhostid); } +boolean_t +spa_trust_config(spa_t *spa) +{ + return (spa->spa_trust_config); +} + +uint64_t +spa_missing_tvds_allowed(spa_t *spa) +{ + return (spa->spa_missing_tvds_allowed); +} + +void +spa_set_missing_tvds(spa_t *spa, uint64_t missing) +{ + spa->spa_missing_tvds = missing; +} + #if defined(_KERNEL) && defined(HAVE_SPL) #include <linux/mod_compat.h> @@ -2338,6 +2359,9 @@ EXPORT_SYMBOL(spa_is_root); EXPORT_SYMBOL(spa_writeable); EXPORT_SYMBOL(spa_mode); EXPORT_SYMBOL(spa_namespace_lock); +EXPORT_SYMBOL(spa_trust_config); +EXPORT_SYMBOL(spa_missing_tvds_allowed); +EXPORT_SYMBOL(spa_set_missing_tvds); /* BEGIN CSTYLED */ module_param(zfs_flags, uint, 0644); diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c index 3654919fc..ad53c0c89 100644 --- a/module/zfs/vdev.c +++ b/module/zfs/vdev.c @@ -74,6 +74,8 @@ unsigned int zfs_checksums_per_second = 20; */ int zfs_scan_ignore_errors = 0; +int vdev_validate_skip = B_FALSE; + /*PRINTFLIKE2*/ void vdev_dbgmsg(vdev_t *vd, const char *fmt, ...) @@ -96,6 +98,57 @@ vdev_dbgmsg(vdev_t *vd, const char *fmt, ...) } } +void +vdev_dbgmsg_print_tree(vdev_t *vd, int indent) +{ + char state[20]; + + if (vd->vdev_ishole || vd->vdev_ops == &vdev_missing_ops) { + zfs_dbgmsg("%*svdev %u: %s", indent, "", vd->vdev_id, + vd->vdev_ops->vdev_op_type); + return; + } + + switch (vd->vdev_state) { + case VDEV_STATE_UNKNOWN: + (void) snprintf(state, sizeof (state), "unknown"); + break; + case VDEV_STATE_CLOSED: + (void) snprintf(state, sizeof (state), "closed"); + break; + case VDEV_STATE_OFFLINE: + (void) snprintf(state, sizeof (state), "offline"); + break; + case VDEV_STATE_REMOVED: + (void) snprintf(state, sizeof (state), "removed"); + break; + case VDEV_STATE_CANT_OPEN: + (void) snprintf(state, sizeof (state), "can't open"); + break; + case VDEV_STATE_FAULTED: + (void) snprintf(state, sizeof (state), "faulted"); + break; + case VDEV_STATE_DEGRADED: + (void) snprintf(state, sizeof (state), "degraded"); + break; + case VDEV_STATE_HEALTHY: + (void) snprintf(state, sizeof (state), "healthy"); + break; + default: + (void) snprintf(state, sizeof (state), "<state %u>", + (uint_t)vd->vdev_state); + } + + zfs_dbgmsg("%*svdev %u: %s%s, guid: %llu, path: %s, %s", indent, + "", vd->vdev_id, vd->vdev_ops->vdev_op_type, + vd->vdev_islog ? " (log)" : "", + (u_longlong_t)vd->vdev_guid, + vd->vdev_path ? vd->vdev_path : "N/A", state); + + for (uint64_t i = 0; i < vd->vdev_children; i++) + vdev_dbgmsg_print_tree(vd->vdev_child[i], indent + 2); +} + /* * Virtual device management. */ @@ -1424,8 +1477,13 @@ vdev_open(vdev_t *vd) vd->vdev_stat.vs_aux != VDEV_AUX_OPEN_FAILED) vd->vdev_removed = B_FALSE; - vdev_set_state(vd, B_TRUE, VDEV_STATE_CANT_OPEN, - vd->vdev_stat.vs_aux); + if (vd->vdev_stat.vs_aux == VDEV_AUX_CHILDREN_OFFLINE) { + vdev_set_state(vd, B_TRUE, VDEV_STATE_OFFLINE, + vd->vdev_stat.vs_aux); + } else { + vdev_set_state(vd, B_TRUE, VDEV_STATE_CANT_OPEN, + vd->vdev_stat.vs_aux); + } return (error); } @@ -1596,29 +1654,29 @@ vdev_open(vdev_t *vd) /* * Called once the vdevs are all opened, this routine validates the label - * contents. This needs to be done before vdev_load() so that we don't + * contents. This needs to be done before vdev_load() so that we don't * inadvertently do repair I/Os to the wrong device. * - * If 'strict' is false ignore the spa guid check. This is necessary because - * if the machine crashed during a re-guid the new guid might have been written - * to all of the vdev labels, but not the cached config. The strict check - * will be performed when the pool is opened again using the mos config. - * * This function will only return failure if one of the vdevs indicates that it * has since been destroyed or exported. This is only possible if * /etc/zfs/zpool.cache was readonly at the time. Otherwise, the vdev state * will be updated but the function will return 0. */ int -vdev_validate(vdev_t *vd, boolean_t strict) +vdev_validate(vdev_t *vd) { spa_t *spa = vd->vdev_spa; nvlist_t *label; - uint64_t guid = 0, top_guid; + uint64_t guid = 0, aux_guid = 0, top_guid; uint64_t state; + nvlist_t *nvl; + uint64_t txg; - for (int c = 0; c < vd->vdev_children; c++) - if (vdev_validate(vd->vdev_child[c], strict) != 0) + if (vdev_validate_skip) + return (0); + + for (uint64_t c = 0; c < vd->vdev_children; c++) + if (vdev_validate(vd->vdev_child[c]) != 0) return (SET_ERROR(EBADF)); /* @@ -1626,115 +1684,276 @@ vdev_validate(vdev_t *vd, boolean_t strict) * any further validation. Otherwise, label I/O will fail and we will * overwrite the previous state. */ - if (vd->vdev_ops->vdev_op_leaf && vdev_readable(vd)) { - uint64_t aux_guid = 0; - nvlist_t *nvl; - uint64_t txg = spa_last_synced_txg(spa) != 0 ? - spa_last_synced_txg(spa) : -1ULL; + if (!vd->vdev_ops->vdev_op_leaf || !vdev_readable(vd)) + return (0); - if ((label = vdev_label_read_config(vd, txg)) == NULL) { - vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, - VDEV_AUX_BAD_LABEL); - vdev_dbgmsg(vd, "vdev_validate: failed reading config"); - return (0); - } + /* + * If we are performing an extreme rewind, we allow for a label that + * was modified at a point after the current txg. + */ + if (spa->spa_extreme_rewind || spa_last_synced_txg(spa) == 0) + txg = UINT64_MAX; + else + txg = spa_last_synced_txg(spa); - /* - * Determine if this vdev has been split off into another - * pool. If so, then refuse to open it. - */ - if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_SPLIT_GUID, - &aux_guid) == 0 && aux_guid == spa_guid(spa)) { - vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, - VDEV_AUX_SPLIT_POOL); - nvlist_free(label); - vdev_dbgmsg(vd, "vdev_validate: vdev split into other " - "pool"); - return (0); - } + if ((label = vdev_label_read_config(vd, txg)) == NULL) { + vdev_set_state(vd, B_TRUE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_BAD_LABEL); + vdev_dbgmsg(vd, "vdev_validate: failed reading config"); + return (0); + } - if (strict && (nvlist_lookup_uint64(label, - ZPOOL_CONFIG_POOL_GUID, &guid) != 0 || - guid != spa_guid(spa))) { - vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, - VDEV_AUX_CORRUPT_DATA); - nvlist_free(label); - vdev_dbgmsg(vd, "vdev_validate: vdev label pool_guid " - "doesn't match config (%llu != %llu)", - (u_longlong_t)guid, - (u_longlong_t)spa_guid(spa)); - return (0); - } + /* + * Determine if this vdev has been split off into another + * pool. If so, then refuse to open it. + */ + if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_SPLIT_GUID, + &aux_guid) == 0 && aux_guid == spa_guid(spa)) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_SPLIT_POOL); + nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: vdev split into other pool"); + return (0); + } - if (nvlist_lookup_nvlist(label, ZPOOL_CONFIG_VDEV_TREE, &nvl) - != 0 || nvlist_lookup_uint64(nvl, ZPOOL_CONFIG_ORIG_GUID, - &aux_guid) != 0) - aux_guid = 0; + if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_POOL_GUID, &guid) != 0) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_CORRUPT_DATA); + nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: '%s' missing from label", + ZPOOL_CONFIG_POOL_GUID); + return (0); + } - /* - * If this vdev just became a top-level vdev because its - * sibling was detached, it will have adopted the parent's - * vdev guid -- but the label may or may not be on disk yet. - * Fortunately, either version of the label will have the - * same top guid, so if we're a top-level vdev, we can - * safely compare to that instead. - * - * If we split this vdev off instead, then we also check the - * original pool's guid. We don't want to consider the vdev - * corrupt if it is partway through a split operation. - */ - if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_GUID, - &guid) != 0 || - nvlist_lookup_uint64(label, ZPOOL_CONFIG_TOP_GUID, - &top_guid) != 0 || - ((vd->vdev_guid != guid && vd->vdev_guid != aux_guid) && - (vd->vdev_guid != top_guid || vd != vd->vdev_top))) { - vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, - VDEV_AUX_CORRUPT_DATA); - nvlist_free(label); - vdev_dbgmsg(vd, "vdev_validate: config guid doesn't " - "match label guid (%llu != %llu)", - (u_longlong_t)vd->vdev_guid, (u_longlong_t)guid); - return (0); + /* + * If config is not trusted then ignore the spa guid check. This is + * necessary because if the machine crashed during a re-guid the new + * guid might have been written to all of the vdev labels, but not the + * cached config. The check will be performed again once we have the + * trusted config from the MOS. + */ + if (spa->spa_trust_config && guid != spa_guid(spa)) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_CORRUPT_DATA); + nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: vdev label pool_guid doesn't " + "match config (%llu != %llu)", (u_longlong_t)guid, + (u_longlong_t)spa_guid(spa)); + return (0); + } + + if (nvlist_lookup_nvlist(label, ZPOOL_CONFIG_VDEV_TREE, &nvl) + != 0 || nvlist_lookup_uint64(nvl, ZPOOL_CONFIG_ORIG_GUID, + &aux_guid) != 0) + aux_guid = 0; + + if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_GUID, &guid) != 0) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_CORRUPT_DATA); + nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: '%s' missing from label", + ZPOOL_CONFIG_GUID); + return (0); + } + + if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_TOP_GUID, &top_guid) + != 0) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_CORRUPT_DATA); + nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: '%s' missing from label", + ZPOOL_CONFIG_TOP_GUID); + return (0); + } + + /* + * If this vdev just became a top-level vdev because its sibling was + * detached, it will have adopted the parent's vdev guid -- but the + * label may or may not be on disk yet. Fortunately, either version + * of the label will have the same top guid, so if we're a top-level + * vdev, we can safely compare to that instead. + * However, if the config comes from a cachefile that failed to update + * after the detach, a top-level vdev will appear as a non top-level + * vdev in the config. Also relax the constraints if we perform an + * extreme rewind. + * + * If we split this vdev off instead, then we also check the + * original pool's guid. We don't want to consider the vdev + * corrupt if it is partway through a split operation. + */ + if (vd->vdev_guid != guid && vd->vdev_guid != aux_guid) { + boolean_t mismatch = B_FALSE; + if (spa->spa_trust_config && !spa->spa_extreme_rewind) { + if (vd != vd->vdev_top || vd->vdev_guid != top_guid) + mismatch = B_TRUE; + } else { + if (vd->vdev_guid != top_guid && + vd->vdev_top->vdev_guid != guid) + mismatch = B_TRUE; } - if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_POOL_STATE, - &state) != 0) { + if (mismatch) { vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, VDEV_AUX_CORRUPT_DATA); nvlist_free(label); - vdev_dbgmsg(vd, "vdev_validate: '%s' missing", - ZPOOL_CONFIG_POOL_STATE); + vdev_dbgmsg(vd, "vdev_validate: config guid " + "doesn't match label guid"); + vdev_dbgmsg(vd, "CONFIG: guid %llu, top_guid %llu", + (u_longlong_t)vd->vdev_guid, + (u_longlong_t)vd->vdev_top->vdev_guid); + vdev_dbgmsg(vd, "LABEL: guid %llu, top_guid %llu, " + "aux_guid %llu", (u_longlong_t)guid, + (u_longlong_t)top_guid, (u_longlong_t)aux_guid); return (0); } + } + if (nvlist_lookup_uint64(label, ZPOOL_CONFIG_POOL_STATE, + &state) != 0) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_CORRUPT_DATA); nvlist_free(label); + vdev_dbgmsg(vd, "vdev_validate: '%s' missing from label", + ZPOOL_CONFIG_POOL_STATE); + return (0); + } - /* - * If this is a verbatim import, no need to check the - * state of the pool. - */ - if (!(spa->spa_import_flags & ZFS_IMPORT_VERBATIM) && - spa_load_state(spa) == SPA_LOAD_OPEN && - state != POOL_STATE_ACTIVE) { - vdev_dbgmsg(vd, "vdev_validate: invalid pool state " - "(%llu) for spa %s", (u_longlong_t)state, - spa->spa_name); - return (SET_ERROR(EBADF)); + nvlist_free(label); + + /* + * If this is a verbatim import, no need to check the + * state of the pool. + */ + if (!(spa->spa_import_flags & ZFS_IMPORT_VERBATIM) && + spa_load_state(spa) == SPA_LOAD_OPEN && + state != POOL_STATE_ACTIVE) { + vdev_dbgmsg(vd, "vdev_validate: invalid pool state (%llu) " + "for spa %s", (u_longlong_t)state, spa->spa_name); + return (SET_ERROR(EBADF)); + } + + /* + * If we were able to open and validate a vdev that was + * previously marked permanently unavailable, clear that state + * now. + */ + if (vd->vdev_not_present) + vd->vdev_not_present = 0; + + return (0); +} + +static void +vdev_copy_path_impl(vdev_t *svd, vdev_t *dvd) +{ + if (svd->vdev_path != NULL && dvd->vdev_path != NULL) { + if (strcmp(svd->vdev_path, dvd->vdev_path) != 0) { + zfs_dbgmsg("vdev_copy_path: vdev %llu: path changed " + "from '%s' to '%s'", (u_longlong_t)dvd->vdev_guid, + dvd->vdev_path, svd->vdev_path); + spa_strfree(dvd->vdev_path); + dvd->vdev_path = spa_strdup(svd->vdev_path); } + } else if (svd->vdev_path != NULL) { + dvd->vdev_path = spa_strdup(svd->vdev_path); + zfs_dbgmsg("vdev_copy_path: vdev %llu: path set to '%s'", + (u_longlong_t)dvd->vdev_guid, dvd->vdev_path); + } +} - /* - * If we were able to open and validate a vdev that was - * previously marked permanently unavailable, clear that state - * now. - */ - if (vd->vdev_not_present) - vd->vdev_not_present = 0; +/* + * Recursively copy vdev paths from one vdev to another. Source and destination + * vdev trees must have same geometry otherwise return error. Intended to copy + * paths from userland config into MOS config. + */ +int +vdev_copy_path_strict(vdev_t *svd, vdev_t *dvd) +{ + if ((svd->vdev_ops == &vdev_missing_ops) || + (svd->vdev_ishole && dvd->vdev_ishole) || + (dvd->vdev_ops == &vdev_indirect_ops)) + return (0); + + if (svd->vdev_ops != dvd->vdev_ops) { + vdev_dbgmsg(svd, "vdev_copy_path: vdev type mismatch: %s != %s", + svd->vdev_ops->vdev_op_type, dvd->vdev_ops->vdev_op_type); + return (SET_ERROR(EINVAL)); + } + + if (svd->vdev_guid != dvd->vdev_guid) { + vdev_dbgmsg(svd, "vdev_copy_path: guids mismatch (%llu != " + "%llu)", (u_longlong_t)svd->vdev_guid, + (u_longlong_t)dvd->vdev_guid); + return (SET_ERROR(EINVAL)); } + if (svd->vdev_children != dvd->vdev_children) { + vdev_dbgmsg(svd, "vdev_copy_path: children count mismatch: " + "%llu != %llu", (u_longlong_t)svd->vdev_children, + (u_longlong_t)dvd->vdev_children); + return (SET_ERROR(EINVAL)); + } + + for (uint64_t i = 0; i < svd->vdev_children; i++) { + int error = vdev_copy_path_strict(svd->vdev_child[i], + dvd->vdev_child[i]); + if (error != 0) + return (error); + } + + if (svd->vdev_ops->vdev_op_leaf) + vdev_copy_path_impl(svd, dvd); + return (0); } +static void +vdev_copy_path_search(vdev_t *stvd, vdev_t *dvd) +{ + ASSERT(stvd->vdev_top == stvd); + ASSERT3U(stvd->vdev_id, ==, dvd->vdev_top->vdev_id); + + for (uint64_t i = 0; i < dvd->vdev_children; i++) { + vdev_copy_path_search(stvd, dvd->vdev_child[i]); + } + + if (!dvd->vdev_ops->vdev_op_leaf || !vdev_is_concrete(dvd)) + return; + + /* + * The idea here is that while a vdev can shift positions within + * a top vdev (when replacing, attaching mirror, etc.) it cannot + * step outside of it. + */ + vdev_t *vd = vdev_lookup_by_guid(stvd, dvd->vdev_guid); + + if (vd == NULL || vd->vdev_ops != dvd->vdev_ops) + return; + + ASSERT(vd->vdev_ops->vdev_op_leaf); + + vdev_copy_path_impl(vd, dvd); +} + +/* + * Recursively copy vdev paths from one root vdev to another. Source and + * destination vdev trees may differ in geometry. For each destination leaf + * vdev, search a vdev with the same guid and top vdev id in the source. + * Intended to copy paths from userland config into MOS config. + */ +void +vdev_copy_path_relaxed(vdev_t *srvd, vdev_t *drvd) +{ + uint64_t children = MIN(srvd->vdev_children, drvd->vdev_children); + ASSERT(srvd->vdev_ops == &vdev_root_ops); + ASSERT(drvd->vdev_ops == &vdev_root_ops); + + for (uint64_t i = 0; i < children; i++) { + vdev_copy_path_search(srvd->vdev_child[i], + drvd->vdev_child[i]); + } +} + /* * Close a virtual device. */ @@ -1828,7 +2047,7 @@ vdev_reopen(vdev_t *vd) !l2arc_vdev_present(vd)) l2arc_add_vdev(spa, vd); } else { - (void) vdev_validate(vd, B_TRUE); + (void) vdev_validate(vd); } /* @@ -3873,6 +4092,19 @@ vdev_set_state(vdev_t *vd, boolean_t isopen, vdev_state_t state, vdev_aux_t aux) vdev_propagate_state(vd->vdev_parent); } +boolean_t +vdev_children_are_offline(vdev_t *vd) +{ + ASSERT(!vd->vdev_ops->vdev_op_leaf); + + for (uint64_t i = 0; i < vd->vdev_children; i++) { + if (vd->vdev_child[i]->vdev_state != VDEV_STATE_OFFLINE) + return (B_FALSE); + } + + return (B_TRUE); +} + /* * Check the vdev configuration to ensure that it's capable of supporting * a root pool. We do not support partial configuration. @@ -3909,34 +4141,6 @@ vdev_is_concrete(vdev_t *vd) } /* - * Load the state from the original vdev tree (ovd) which - * we've retrieved from the MOS config object. If the original - * vdev was offline or faulted then we transfer that state to the - * device in the current vdev tree (nvd). - */ -void -vdev_load_log_state(vdev_t *nvd, vdev_t *ovd) -{ - ASSERT(nvd->vdev_top->vdev_islog); - ASSERT(spa_config_held(nvd->vdev_spa, - SCL_STATE_ALL, RW_WRITER) == SCL_STATE_ALL); - ASSERT3U(nvd->vdev_guid, ==, ovd->vdev_guid); - - for (int c = 0; c < nvd->vdev_children; c++) - vdev_load_log_state(nvd->vdev_child[c], ovd->vdev_child[c]); - - if (nvd->vdev_ops->vdev_op_leaf) { - /* - * Restore the persistent vdev state - */ - nvd->vdev_offline = ovd->vdev_offline; - nvd->vdev_faulted = ovd->vdev_faulted; - nvd->vdev_degraded = ovd->vdev_degraded; - nvd->vdev_removed = ovd->vdev_removed; - } -} - -/* * Determine if a log device has valid content. If the vdev was * removed or faulted in the MOS config then we know that * the content on the log device has already been written to the pool. @@ -4051,5 +4255,9 @@ module_param(zfs_checksums_per_second, uint, 0644); module_param(zfs_scan_ignore_errors, int, 0644); MODULE_PARM_DESC(zfs_scan_ignore_errors, "Ignore errors during resilver/scrub"); + +module_param(vdev_validate_skip, int, 0644); +MODULE_PARM_DESC(vdev_validate_skip, + "Bypass vdev_validate()"); /* END CSTYLED */ #endif diff --git a/module/zfs/vdev_label.c b/module/zfs/vdev_label.c index ad334fe8c..85d133a5a 100644 --- a/module/zfs/vdev_label.c +++ b/module/zfs/vdev_label.c @@ -412,7 +412,7 @@ vdev_config_generate(spa_t *spa, vdev_t *vd, boolean_t getstats, fnvlist_add_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK, vd->vdev_wholedisk); - if (vd->vdev_not_present) + if (vd->vdev_not_present && !(flags & VDEV_CONFIG_MISSING)) fnvlist_add_uint64(nv, ZPOOL_CONFIG_NOT_PRESENT, 1); if (vd->vdev_isspare) @@ -1209,6 +1209,11 @@ vdev_uberblock_load(vdev_t *rvd, uberblock_t *ub, nvlist_t **config) "txg %llu", spa->spa_name, (u_longlong_t)ub->ub_txg); *config = vdev_label_read_config(cb.ubl_vd, ub->ub_txg); + if (*config == NULL && spa->spa_extreme_rewind) { + vdev_dbgmsg(cb.ubl_vd, "failed to read label config. " + "Trying again without txg restrictions."); + *config = vdev_label_read_config(cb.ubl_vd, UINT64_MAX); + } if (*config == NULL) { vdev_dbgmsg(cb.ubl_vd, "failed to read label config"); } diff --git a/module/zfs/vdev_mirror.c b/module/zfs/vdev_mirror.c index 4b01f317b..1c591cd64 100644 --- a/module/zfs/vdev_mirror.c +++ b/module/zfs/vdev_mirror.c @@ -251,9 +251,33 @@ vdev_mirror_map_init(zio_t *zio) if (vd == NULL) { dva_t *dva = zio->io_bp->blk_dva; spa_t *spa = zio->io_spa; + dva_t dva_copy[SPA_DVAS_PER_BP]; - mm = vdev_mirror_map_alloc(BP_GET_NDVAS(zio->io_bp), B_FALSE, - B_TRUE); + c = BP_GET_NDVAS(zio->io_bp); + + /* + * If we do not trust the pool config, some DVAs might be + * invalid or point to vdevs that do not exist. We skip them. + */ + if (!spa_trust_config(spa)) { + ASSERT3U(zio->io_type, ==, ZIO_TYPE_READ); + int j = 0; + for (int i = 0; i < c; i++) { + if (zfs_dva_valid(spa, &dva[i], zio->io_bp)) + dva_copy[j++] = dva[i]; + } + if (j == 0) { + zio->io_vsd = NULL; + zio->io_error = ENXIO; + return (NULL); + } + if (j < c) { + dva = dva_copy; + c = j; + } + } + + mm = vdev_mirror_map_alloc(c, B_FALSE, B_TRUE); for (c = 0; c < mm->mm_children; c++) { mc = &mm->mm_child[c]; @@ -305,7 +329,10 @@ vdev_mirror_open(vdev_t *vd, uint64_t *asize, uint64_t *max_asize, } if (numerrors == vd->vdev_children) { - vd->vdev_stat.vs_aux = VDEV_AUX_NO_REPLICAS; + if (vdev_children_are_offline(vd)) + vd->vdev_stat.vs_aux = VDEV_AUX_CHILDREN_OFFLINE; + else + vd->vdev_stat.vs_aux = VDEV_AUX_NO_REPLICAS; return (lasterror); } @@ -485,6 +512,13 @@ vdev_mirror_io_start(zio_t *zio) mm = vdev_mirror_map_init(zio); + if (mm == NULL) { + ASSERT(!spa_trust_config(zio->io_spa)); + ASSERT(zio->io_type == ZIO_TYPE_READ); + zio_execute(zio); + return; + } + if (zio->io_type == ZIO_TYPE_READ) { if (zio->io_bp != NULL && (zio->io_flags & ZIO_FLAG_SCRUB) && !mm->mm_replacing) { @@ -558,6 +592,9 @@ vdev_mirror_io_done(zio_t *zio) int good_copies = 0; int unexpected_errors = 0; + if (mm == NULL) + return; + for (c = 0; c < mm->mm_children; c++) { mc = &mm->mm_child[c]; @@ -677,13 +714,19 @@ vdev_mirror_io_done(zio_t *zio) static void vdev_mirror_state_change(vdev_t *vd, int faulted, int degraded) { - if (faulted == vd->vdev_children) - vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, - VDEV_AUX_NO_REPLICAS); - else if (degraded + faulted != 0) + if (faulted == vd->vdev_children) { + if (vdev_children_are_offline(vd)) { + vdev_set_state(vd, B_FALSE, VDEV_STATE_OFFLINE, + VDEV_AUX_CHILDREN_OFFLINE); + } else { + vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, + VDEV_AUX_NO_REPLICAS); + } + } else if (degraded + faulted != 0) { vdev_set_state(vd, B_FALSE, VDEV_STATE_DEGRADED, VDEV_AUX_NONE); - else + } else { vdev_set_state(vd, B_FALSE, VDEV_STATE_HEALTHY, VDEV_AUX_NONE); + } } vdev_ops_t vdev_mirror_ops = { diff --git a/module/zfs/vdev_root.c b/module/zfs/vdev_root.c index 8ac9ce187..9f86cbfa4 100644 --- a/module/zfs/vdev_root.c +++ b/module/zfs/vdev_root.c @@ -37,6 +37,23 @@ * Virtual device vector for the pool's root vdev. */ +static uint64_t +vdev_root_core_tvds(vdev_t *vd) +{ + uint64_t tvds = 0; + + for (uint64_t c = 0; c < vd->vdev_children; c++) { + vdev_t *cvd = vd->vdev_child[c]; + + if (!cvd->vdev_ishole && !cvd->vdev_islog && + cvd->vdev_ops != &vdev_indirect_ops) { + tvds++; + } + } + + return (tvds); +} + /* * We should be able to tolerate one failure with absolutely no damage * to our metadata. Two failures will take out space maps, a bunch of @@ -46,17 +63,28 @@ * probably fine. Adding bean counters during alloc/free can make this * future guesswork more accurate. */ -static int -too_many_errors(vdev_t *vd, int numerrors) +static boolean_t +too_many_errors(vdev_t *vd, uint64_t numerrors) { - ASSERT3U(numerrors, <=, vd->vdev_children); - return (numerrors > 0); + uint64_t tvds; + + if (numerrors == 0) + return (B_FALSE); + + tvds = vdev_root_core_tvds(vd); + ASSERT3U(numerrors, <=, tvds); + + if (numerrors == tvds) + return (B_TRUE); + + return (numerrors > spa_missing_tvds_allowed(vd->vdev_spa)); } static int vdev_root_open(vdev_t *vd, uint64_t *asize, uint64_t *max_asize, uint64_t *ashift) { + spa_t *spa = vd->vdev_spa; int lasterror = 0; int numerrors = 0; @@ -76,6 +104,9 @@ vdev_root_open(vdev_t *vd, uint64_t *asize, uint64_t *max_asize, } } + if (spa_load_state(spa) != SPA_LOAD_NONE) + spa_set_missing_tvds(spa, numerrors); + if (too_many_errors(vd, numerrors)) { vd->vdev_stat.vs_aux = VDEV_AUX_NO_REPLICAS; return (lasterror); @@ -101,7 +132,7 @@ vdev_root_state_change(vdev_t *vd, int faulted, int degraded) if (too_many_errors(vd, faulted)) { vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN, VDEV_AUX_NO_REPLICAS); - } else if (degraded) { + } else if (degraded || faulted) { vdev_set_state(vd, B_FALSE, VDEV_STATE_DEGRADED, VDEV_AUX_NONE); } else { vdev_set_state(vd, B_FALSE, VDEV_STATE_HEALTHY, VDEV_AUX_NONE); diff --git a/module/zfs/zio.c b/module/zfs/zio.c index 6822505f1..81ae65c31 100644 --- a/module/zfs/zio.c +++ b/module/zfs/zio.c @@ -879,6 +879,13 @@ zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp) } /* + * Do not verify individual DVAs if the config is not trusted. This + * will be done once the zio is executed in vdev_mirror_map_alloc. + */ + if (!spa->spa_trust_config) + return; + + /* * Pool-specific checks. * * Note: it would be nice to verify that the blk_birth and @@ -928,6 +935,36 @@ zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp) } } +boolean_t +zfs_dva_valid(spa_t *spa, const dva_t *dva, const blkptr_t *bp) +{ + uint64_t vdevid = DVA_GET_VDEV(dva); + + if (vdevid >= spa->spa_root_vdev->vdev_children) + return (B_FALSE); + + vdev_t *vd = spa->spa_root_vdev->vdev_child[vdevid]; + if (vd == NULL) + return (B_FALSE); + + if (vd->vdev_ops == &vdev_hole_ops) + return (B_FALSE); + + if (vd->vdev_ops == &vdev_missing_ops) { + return (B_FALSE); + } + + uint64_t offset = DVA_GET_OFFSET(dva); + uint64_t asize = DVA_GET_ASIZE(dva); + + if (BP_IS_GANG(bp)) + asize = vdev_psize_to_asize(vd, SPA_GANGBLOCKSIZE); + if (offset + asize > vd->vdev_asize) + return (B_FALSE); + + return (B_TRUE); +} + zio_t * zio_read(zio_t *pio, spa_t *spa, const blkptr_t *bp, abd_t *data, uint64_t size, zio_done_func_t *done, void *private, @@ -3473,14 +3510,18 @@ zio_vdev_io_start(zio_t *zio) } ASSERT3P(zio->io_logical, !=, zio); - if (zio->io_type == ZIO_TYPE_WRITE && zio->io_vd->vdev_removing) { + if (zio->io_type == ZIO_TYPE_WRITE) { + ASSERT(spa->spa_trust_config); + /* * Note: the code can handle other kinds of writes, * but we don't expect them. */ - ASSERT(zio->io_flags & - (ZIO_FLAG_PHYSICAL | ZIO_FLAG_SELF_HEAL | - ZIO_FLAG_RESILVER | ZIO_FLAG_INDUCE_DAMAGE)); + if (zio->io_vd->vdev_removing) { + ASSERT(zio->io_flags & + (ZIO_FLAG_PHYSICAL | ZIO_FLAG_SELF_HEAL | + ZIO_FLAG_RESILVER | ZIO_FLAG_INDUCE_DAMAGE)); + } } align = 1ULL << vd->vdev_top->vdev_ashift; diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run index eecac8f6e..0260eb884 100644 --- a/tests/runfiles/linux.run +++ b/tests/runfiles/linux.run @@ -364,11 +364,22 @@ tests = ['zpool_import_001_pos', 'zpool_import_002_pos', 'zpool_import_012_pos', 'zpool_import_013_neg', 'zpool_import_014_pos', 'zpool_import_015_pos', 'zpool_import_features_001_pos', 'zpool_import_features_002_neg', - 'zpool_import_features_003_pos','zpool_import_missing_001_pos', + 'zpool_import_features_003_pos', 'zpool_import_missing_001_pos', 'zpool_import_missing_002_pos', 'zpool_import_rename_001_pos', 'zpool_import_all_001_pos', 'zpool_import_encrypted', 'zpool_import_encrypted_load', - 'zpool_import_errata3'] + 'zpool_import_errata3', + 'import_cache_device_added', + 'import_cache_device_removed', + 'import_cache_device_replaced', + 'import_cache_mirror_attached', + 'import_cache_mirror_detached', + 'import_cache_shared_device', + 'import_devices_missing', + 'import_paths_changed', + 'import_rewind_config_changed', + 'import_rewind_device_replaced'] + tags = ['functional', 'cli_root', 'zpool_import'] [tests/functional/cli_root/zpool_labelclear] diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am index 8aa34f33c..97a15a20d 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/Makefile.am @@ -2,6 +2,17 @@ pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/cli_root/zpool_impo dist_pkgdata_SCRIPTS = \ setup.ksh \ cleanup.ksh \ + zpool_import.kshlib \ + import_cache_device_added.ksh \ + import_cache_device_removed.ksh \ + import_cache_device_replaced.ksh \ + import_cache_mirror_attached.ksh \ + import_cache_mirror_detached.ksh \ + import_cache_shared_device.ksh \ + import_devices_missing.ksh \ + import_paths_changed.ksh \ + import_rewind_config_changed.ksh \ + import_rewind_device_replaced.ksh \ zpool_import_001_pos.ksh \ zpool_import_002_pos.ksh \ zpool_import_003_pos.ksh \ diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_added.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_added.ksh new file mode 100755 index 000000000..bda6b891b --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_added.ksh @@ -0,0 +1,76 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable using an outdated cachefile that is unaware +# that one or two top-level vdevs were added. +# +# STRATEGY: +# 1. Create a pool with some devices and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Add a device/mirror/raid to the pool. +# 4. Export the pool. +# 5. Verify that we can import the pool using the backed-up cachefile. +# + +verify_runnable "global" + +log_onexit cleanup + +function test_add_vdevs +{ + typeset poolcreate="$1" + typeset addvdevs="$2" + typeset poolcheck="$3" + + log_note "$0: pool '$poolcreate', add $addvdevs." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $poolcreate + + log_must cp $CPATH $CPATHBKP + + log_must zpool add -f $TESTPOOL1 $addvdevs + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcheck" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + + log_note "" +} + +test_add_vdevs "$VDEV0" "$VDEV1" "$VDEV0 $VDEV1" +test_add_vdevs "$VDEV0 $VDEV1" "$VDEV2" "$VDEV0 $VDEV1 $VDEV2" +test_add_vdevs "$VDEV0" "$VDEV1 $VDEV2" "$VDEV0 $VDEV1 $VDEV2" +test_add_vdevs "$VDEV0" "mirror $VDEV1 $VDEV2" \ + "$VDEV0 mirror $VDEV1 $VDEV2" +test_add_vdevs "mirror $VDEV0 $VDEV1" "mirror $VDEV2 $VDEV3" \ + "mirror $VDEV0 $VDEV1 mirror $VDEV2 $VDEV3" +test_add_vdevs "$VDEV0" "raidz $VDEV1 $VDEV2 $VDEV3" \ + "$VDEV0 raidz $VDEV1 $VDEV2 $VDEV3" +test_add_vdevs "$VDEV0" "log $VDEV1" "$VDEV0 log $VDEV1" +test_add_vdevs "$VDEV0 log $VDEV1" "$VDEV2" "$VDEV0 $VDEV2 log $VDEV1" +test_add_vdevs "$VDEV0" "$VDEV1 log $VDEV2" "$VDEV0 $VDEV1 log $VDEV2" + +log_pass "zpool import -c cachefile_unaware_of_add passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_removed.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_removed.ksh new file mode 100755 index 000000000..1d878b7a2 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_removed.ksh @@ -0,0 +1,145 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable using an outdated cachefile that is unaware +# that one or more vdevs were removed. +# +# STRATEGY: +# 1. Create a pool with some devices and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Remove device(s) from the pool and remove them. +# 4. (Optionally) Add device(s) to pool. +# 5. Export the pool. +# 6. Verify that we can import the pool using the backed-up cachefile. +# + +verify_runnable "global" + +function custom_cleanup +{ + cleanup +} + +log_onexit custom_cleanup + +function test_remove_vdev +{ + typeset poolcreate="$1" + typeset removevdev="$2" + typeset poolcheck="$3" + + log_note "$0: pool '$poolcreate', remove $2." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $poolcreate + + log_must cp $CPATH $CPATHBKP + + log_must zpool remove $TESTPOOL1 $removevdev + log_must wait_for_pool_config $TESTPOOL1 "$poolcheck" + log_must rm $removevdev + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcheck" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + log_must mkfile $FILE_SIZE $removevdev + + log_note "" +} + +# +# We have to remove top-level non-log vdevs one by one, else there is a high +# chance pool will report busy and command will fail for the second vdev. +# +function test_remove_two_vdevs +{ + log_note "$0." + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 \ + $VDEV0 $VDEV1 $VDEV2 $VDEV3 $VDEV4 + + log_must cp $CPATH $CPATHBKP + + log_must zpool remove $TESTPOOL1 $VDEV4 + log_must wait_for_pool_config $TESTPOOL1 \ + "$VDEV0 $VDEV1 $VDEV2 $VDEV3" + log_must zpool remove $TESTPOOL1 $VDEV3 + log_must wait_for_pool_config $TESTPOOL1 "$VDEV0 $VDEV1 $VDEV2" + log_must rm $VDEV3 $VDEV4 + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$VDEV0 $VDEV1 $VDEV2" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + log_must mkfile $FILE_SIZE $VDEV3 $VDEV4 + + log_note "" +} + +# +# We want to test the case where a whole created by a log device is filled +# by a regular device +# +function test_remove_log_then_add_vdev +{ + log_note "$0." + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 \ + $VDEV0 $VDEV1 $VDEV2 log $VDEV3 + + log_must cp $CPATH $CPATHBKP + + log_must zpool remove $TESTPOOL1 $VDEV1 + log_must wait_for_pool_config $TESTPOOL1 "$VDEV0 $VDEV2 log $VDEV3" + log_must zpool remove $TESTPOOL1 $VDEV3 + log_must check_pool_config $TESTPOOL1 "$VDEV0 $VDEV2" + log_must rm $VDEV1 $VDEV3 + log_must zpool add $TESTPOOL1 $VDEV4 + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$VDEV0 $VDEV2 $VDEV4" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + log_must mkfile $FILE_SIZE $VDEV1 $VDEV3 + + log_note "" +} + +test_remove_vdev "$VDEV0 $VDEV1 $VDEV2" "$VDEV2" "$VDEV0 $VDEV1" +test_remove_vdev "$VDEV0 $VDEV1 $VDEV2" "$VDEV1" "$VDEV0 $VDEV2" +test_remove_vdev "$VDEV0 log $VDEV1" "$VDEV1" "$VDEV0" +test_remove_vdev "$VDEV0 log $VDEV1 $VDEV2" "$VDEV1 $VDEV2" "$VDEV0" +test_remove_vdev "$VDEV0 $VDEV1 $VDEV2 log $VDEV3" "$VDEV2" \ + "$VDEV0 $VDEV1 log $VDEV3" +test_remove_two_vdevs +test_remove_log_then_add_vdev + +log_pass "zpool import -c cachefile_unaware_of_remove passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_replaced.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_replaced.ksh new file mode 100755 index 000000000..f2888a5bb --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_device_replaced.ksh @@ -0,0 +1,166 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable using an outdated cachefile that is unaware +# of a zpool replace operation at different stages in time. +# +# STRATEGY: +# 1. Create a pool with some devices and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Initiate device replacement, backup cachefile again and export pool. +# Special care must be taken so that resilvering doesn't complete +# before we exported the pool. +# 4. Verify that we can import the pool using the first cachefile backup. +# (Test 1. cachefile: pre-replace, pool: resilvering) +# 5. Wait for the resilvering to finish and export the pool. +# 6. Verify that we can import the pool using the first cachefile backup. +# (Test 2. cachefile: pre-replace, pool: post-replace) +# 7. Export the pool. +# 8. Verify that we can import the pool using the second cachefile backup. +# (Test 3. cachefile: resilvering, pool: post-replace) +# +# STRATEGY TO SLOW DOWN RESILVERING: +# 1. Reduce zfs_txg_timeout, which controls how long can we resilver for +# each sync. +# 2. Add data to pool +# 3. Re-import the pool so that data isn't cached +# 4. Use zinject to slow down device I/O +# 5. Trigger the resilvering +# 6. Use spa freeze to stop writing to the pool. +# 7. Clear zinject events (needed to export the pool) +# 8. Export the pool +# + +verify_runnable "global" + +ZFS_TXG_TIMEOUT="" + +function custom_cleanup +{ + # Revert zfs_txg_timeout to defaults + [[ -n ZFS_TXG_TIMEOUT ]] && + log_must set_zfs_txg_timeout $ZFS_TXG_TIMEOUT + + zinject -c all + cleanup +} + +log_onexit custom_cleanup + +function test_replacing_vdevs +{ + typeset poolcreate="$1" + typeset replacevdev="$2" + typeset replaceby="$3" + typeset poolfinalstate="$4" + typeset zinjectdevices="$5" + typeset earlyremove="$6" + typeset writedata="$7" + + log_note "$0: pool '$poolcreate', replace $replacevdev by $replaceby." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $poolcreate + + # Cachefile: pool in pre-replace state + log_must cp $CPATH $CPATHBKP + + # Steps to insure resilvering happens very slowly. + log_must write_some_data $TESTPOOL1 $writedata + log_must zpool export $TESTPOOL1 + log_must cp $CPATHBKP $CPATH + log_must zpool import -c $CPATH -o cachefile=$CPATH $TESTPOOL1 + typeset device + for device in $zinjectdevices ; do + log_must zinject -d $device -D 200:1 $TESTPOOL1 > /dev/null + done + log_must zpool replace $TESTPOOL1 $replacevdev $replaceby + + # Cachefile: pool in resilvering state + log_must cp $CPATH $CPATHBKP2 + + # We must disable zinject in order to export the pool, so we freeze + # it first to prevent writing out subsequent resilvering progress. + log_must zpool freeze $TESTPOOL1 + # Confirm pool is still replacing + log_must pool_is_replacing $TESTPOOL1 + log_must zinject -c all > /dev/null + log_must zpool export $TESTPOOL1 + + ( $earlyremove ) && log_must rm $replacevdev + + ############################################################ + # Test 1. Cachefile: pre-replace, pool: resilvering + ############################################################ + log_must cp $CPATHBKP $CPATH + log_must zpool import -c $CPATH $TESTPOOL1 + + # Wait for resilvering to finish + log_must wait_for_pool_config $TESTPOOL1 "$poolfinalstate" + log_must zpool export $TESTPOOL1 + + ( ! $earlyremove ) && log_must rm $replacevdev + + ############################################################ + # Test 2. Cachefile: pre-replace, pool: post-replace + ############################################################ + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolfinalstate" + log_must zpool export $TESTPOOL1 + + ############################################################ + # Test 3. Cachefile: resilvering, pool: post-replace + ############################################################ + log_must zpool import -c $CPATHBKP2 $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolfinalstate" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP $CPATHBKP2 + log_must mkfile $FILE_SIZE $replacevdev + + log_note "" +} + +# We set zfs_txg_timeout to 1 to reduce resilvering time at each sync. +ZFS_TXG_TIMEOUT=$(get_zfs_txg_timeout) +set_zfs_txg_timeout 1 + +test_replacing_vdevs "$VDEV0 $VDEV1" \ + "$VDEV1" "$VDEV2" \ + "$VDEV0 $VDEV2" \ + "$VDEV0 $VDEV1" \ + false 20 + +test_replacing_vdevs "mirror $VDEV0 $VDEV1" \ + "$VDEV1" "$VDEV2" \ + "mirror $VDEV0 $VDEV2" \ + "$VDEV0 $VDEV1" \ + true 10 + +test_replacing_vdevs "raidz $VDEV0 $VDEV1 $VDEV2" \ + "$VDEV1" "$VDEV3" \ + "raidz $VDEV0 $VDEV3 $VDEV2" \ + "$VDEV0 $VDEV1 $VDEV2" \ + true 20 + +set_zfs_txg_timeout $ZFS_TXG_TIMEOUT + +log_pass "zpool import -c cachefile_unaware_of_replace passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_attached.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_attached.ksh new file mode 100755 index 000000000..987b745b9 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_attached.ksh @@ -0,0 +1,72 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable using an outdated cachefile that misses a +# mirror that was attached. +# +# STRATEGY: +# 1. Create a pool with some devices and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Attach a mirror to one of the devices in the pool. +# 4. Export the pool. +# 5. Verify that we can import the pool using the backed-up cachefile. +# + +verify_runnable "global" + +log_onexit cleanup + +function test_attach_vdev +{ + typeset poolcreate="$1" + typeset attachto="$2" + typeset attachvdev="$3" + typeset poolcheck="$4" + + log_note "$0: pool '$poolcreate', attach $attachvdev to $attachto." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $poolcreate + + log_must cp $CPATH $CPATHBKP + + log_must zpool attach $TESTPOOL1 $attachto $attachvdev + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcheck" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + + log_note "" +} + +test_attach_vdev "$VDEV0" "$VDEV0" "$VDEV4" "mirror $VDEV0 $VDEV4" +test_attach_vdev "$VDEV0 $VDEV1" "$VDEV1" "$VDEV4" \ + "$VDEV0 mirror $VDEV1 $VDEV4" +test_attach_vdev "mirror $VDEV0 $VDEV1" "$VDEV0" "$VDEV4" \ + "mirror $VDEV0 $VDEV1 $VDEV4" +test_attach_vdev "$VDEV0 log $VDEV1" "$VDEV1" "$VDEV4" \ + "$VDEV0 log mirror $VDEV1 $VDEV4" + +log_pass "zpool import -c cachefile_unaware_of_attach passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_detached.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_detached.ksh new file mode 100755 index 000000000..85ec51673 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_mirror_detached.ksh @@ -0,0 +1,70 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable using an outdated cachefile that is unaware +# that a mirror was detached. +# +# STRATEGY: +# 1. Create a pool with some devices mirrored and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Detach a mirror from the pool. +# 4. Export the pool. +# 5. Verify that we can import the pool using the backed-up cachefile. +# + +verify_runnable "global" + +log_onexit cleanup + +function test_detach_vdev +{ + typeset poolcreate="$1" + typeset poolcheck="$2" + + log_note "$0: pool '$poolcreate', detach $VDEV4." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $poolcreate + + log_must cp $CPATH $CPATHBKP + + log_must zpool detach $TESTPOOL1 $VDEV4 + log_must rm -f $VDEV4 + + log_must zpool export $TESTPOOL1 + + log_must zpool import -c $CPATHBKP $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcheck" + + # Cleanup + log_must zpool destroy $TESTPOOL1 + log_must rm -f $CPATH $CPATHBKP + log_must mkfile $FILE_SIZE $VDEV4 + + log_note "" +} + +test_detach_vdev "mirror $VDEV0 $VDEV4" "$VDEV0" +test_detach_vdev "mirror $VDEV0 $VDEV4 mirror $VDEV1 $VDEV2" \ + "$VDEV0 mirror $VDEV1 $VDEV2" +test_detach_vdev "mirror $VDEV0 $VDEV1 $VDEV4" "mirror $VDEV0 $VDEV1" +test_detach_vdev "$VDEV0 log mirror $VDEV1 $VDEV4" "$VDEV0 log $VDEV1" + +log_pass "zpool import -c cachefile_unaware_of_detach passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_shared_device.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_shared_device.ksh new file mode 100755 index 000000000..66225c11b --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_cache_shared_device.ksh @@ -0,0 +1,113 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should not try to write to a device that doesn't belong to it +# anymore, even if the device is in its cachefile. +# +# STRATEGY: +# 1. Create pool1 with some devices and an alternate cachefile. +# 2. Backup the cachefile. +# 3. Export pool1. +# 4. Create pool2 using a device that belongs to pool1. +# 5. Export pool2. +# 6. Compute checksum of the shared device. +# 7. Import pool1 and write some data to it. +# 8. Verify that the checksum of the shared device hasn't changed. +# + +verify_runnable "global" + +function custom_cleanup +{ + destroy_pool $TESTPOOL2 + cleanup +} + +log_onexit custom_cleanup + +function dev_checksum +{ + typeset dev="$1" + typeset checksum + + log_note "Compute checksum of '$dev'" + + checksum=$(md5sum $dev) + if [[ $? -ne 0 ]]; then + log_fail "Failed to compute checksum of '$dev'" + return 1 + fi + + echo "$checksum" + return 0 +} + +function test_shared_device +{ + typeset pool1="$1" + typeset pool2="$2" + typeset sharedvdev="$3" + typeset importflags="${4:-}" + + log_note "$0: pool1 '$pool1', pool2 '$pool2' takes $sharedvdev." + + log_must zpool create -o cachefile=$CPATH $TESTPOOL1 $pool1 + + log_must cp $CPATH $CPATHBKP + + log_must zpool export $TESTPOOL1 + + log_must zpool create -f $TESTPOOL2 $pool2 + + log_must zpool export $TESTPOOL2 + + typeset checksum1=$(dev_checksum $sharedvdev) + + log_must zpool import -c $CPATHBKP $importflags $TESTPOOL1 + + log_must write_some_data $TESTPOOL1 2 + + log_must zpool destroy $TESTPOOL1 + + typeset checksum2=$(dev_checksum $sharedvdev) + + if [[ $checksum1 == $checksum2 ]]; then + log_pos "Device hasn't been modified by original pool" + else + log_fail "Device has been modified by original pool." \ + "Checksum mismatch: $checksum1 != $checksum2." + fi + + # Cleanup + log_must zpool import -d $DEVICE_DIR $TESTPOOL2 + log_must zpool destroy $TESTPOOL2 + log_must rm -f $CPATH $CPATHBKP + + log_note "" +} + +test_shared_device "mirror $VDEV0 $VDEV1" "mirror $VDEV1 $VDEV2" "$VDEV1" +test_shared_device "mirror $VDEV0 $VDEV1 $VDEV2" "mirror $VDEV2 $VDEV3" \ + "$VDEV2" +test_shared_device "raidz $VDEV0 $VDEV1 $VDEV2" "$VDEV2" "$VDEV2" +test_shared_device "$VDEV0 log $VDEV1" "$VDEV2 log $VDEV1" "$VDEV1" "-m" + +log_pass "Pool doesn't write to a device it doesn't own anymore." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_devices_missing.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_devices_missing.ksh new file mode 100755 index 000000000..74b736aef --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_devices_missing.ksh @@ -0,0 +1,122 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable when up to 2 top-level devices are missing. +# +# STRATEGY: +# 1. Create a pool. +# 2. Write some data to the pool and checksum it. +# 3. Add one or more devices. +# 4. Write more data to the pool and checksum it. +# 5. Export the pool. +# 6. Move added devices out of the devices directory. +# 7. Import the pool with missing devices. +# 8. Verify that the first batch of data is intact. +# 9. Verify that accessing the second batch of data doesn't suspend pool. +# 10. Export the pool, move back missing devices, Re-import the pool. +# 11. Verify that all the data is intact. +# + +verify_runnable "global" + +function custom_cleanup +{ + log_must set_spa_load_verify_metadata 1 + log_must set_spa_load_verify_data 1 + log_must set_zfs_max_missing_tvds 0 + log_must rm -rf $BACKUP_DEVICE_DIR + # Highly damaged pools may fail to be destroyed, so we export them. + poolexists $TESTPOOL1 && log_must zpool export $TESTPOOL1 + cleanup +} + +log_onexit custom_cleanup + +function test_devices_missing +{ + typeset poolcreate="$1" + typeset addvdevs="$2" + typeset missingvdevs="$3" + typeset -i missingtvds="$4" + + log_note "$0: pool '$poolcreate', adding $addvdevs, then" \ + "moving away $missingvdevs." + + log_must zpool create $TESTPOOL1 $poolcreate + + log_must generate_data $TESTPOOL1 $MD5FILE "first" + + log_must zpool add $TESTPOOL1 $addvdevs + + log_must generate_data $TESTPOOL1 $MD5FILE2 "second" + + log_must zpool export $TESTPOOL1 + + log_must mv $missingvdevs $BACKUP_DEVICE_DIR + + # Tell zfs that it is ok to import a pool with missing top-level vdevs + log_must set_zfs_max_missing_tvds $missingtvds + # Missing devices means that data or metadata may be corrupted. + (( missingtvds > 1 )) && log_must set_spa_load_verify_metadata 0 + log_must set_spa_load_verify_data 0 + log_must zpool import -o readonly=on -d $DEVICE_DIR $TESTPOOL1 + + log_must verify_data_md5sums $MD5FILE + + log_note "Try reading second batch of data, make sure pool doesn't" \ + "get suspended." + verify_data_md5sums $MD5FILE >/dev/null 2>&1 + + log_must zpool export $TESTPOOL1 + + typeset newpaths=$(echo "$missingvdevs" | \ + sed "s:$DEVICE_DIR:$BACKUP_DEVICE_DIR:g") + log_must mv $newpaths $DEVICE_DIR + log_must set_spa_load_verify_metadata 1 + log_must set_spa_load_verify_data 1 + log_must set_zfs_max_missing_tvds 0 + log_must zpool import -d $DEVICE_DIR $TESTPOOL1 + + log_must verify_data_md5sums $MD5FILE + log_must verify_data_md5sums $MD5FILE2 + + # Cleanup + log_must zpool destroy $TESTPOOL1 + + log_note "" +} + +log_must mkdir -p $BACKUP_DEVICE_DIR + +test_devices_missing "$VDEV0" "$VDEV1" "$VDEV1" 1 +test_devices_missing "$VDEV0" "$VDEV1 $VDEV2" "$VDEV1" 1 +test_devices_missing "mirror $VDEV0 $VDEV1" "mirror $VDEV2 $VDEV3" \ + "$VDEV2 $VDEV3" 1 +test_devices_missing "$VDEV0 log $VDEV1" "$VDEV2" "$VDEV2" 1 + +# +# Note that we are testing for 2 non-consecutive missing devices. +# Missing consecutive devices results in missing metadata. Because of +# Missing metadata can cause the root dataset to fail to mount. +# +test_devices_missing "$VDEV0" "$VDEV1 $VDEV2 $VDEV3" "$VDEV1 $VDEV3" 2 + +log_pass "zpool import succeeded with missing devices." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_paths_changed.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_paths_changed.ksh new file mode 100755 index 000000000..457eb6a14 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_paths_changed.ksh @@ -0,0 +1,98 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# A pool should be importable even if device paths have changed. +# +# STRATEGY: +# 1. Create a pool. +# 2. Export the pool. +# 3. Change the paths of some of the devices. +# 4. Verify that we can import the pool in a healthy state. +# + +verify_runnable "global" + +log_onexit cleanup + +function test_new_paths +{ + typeset poolcreate="$1" + typeset pathstochange="$2" + + log_note "$0: pool '$poolcreate', changing paths of $pathstochange." + + log_must zpool create $TESTPOOL1 $poolcreate + + log_must zpool export $TESTPOOL1 + + for dev in $pathstochange; do + log_must mv $dev "${dev}_new" + done + + log_must zpool import -d $DEVICE_DIR $TESTPOOL1 + log_must check_pool_healthy $TESTPOOL1 + + # Cleanup + log_must zpool destroy $TESTPOOL1 + for dev in $pathstochange; do + log_must mv "${dev}_new" $dev + done + + log_note "" +} + +function test_swap_paths +{ + typeset poolcreate="$1" + typeset pathtoswap1="$2" + typeset pathtoswap2="$3" + + log_note "$0: pool '$poolcreate', swapping paths of $pathtoswap1" \ + "and $pathtoswap2." + + log_must zpool create $TESTPOOL1 $poolcreate + + log_must zpool export $TESTPOOL1 + + log_must mv $pathtoswap2 "$pathtoswap2.tmp" + log_must mv $pathtoswap1 "$pathtoswap2" + log_must mv "$pathtoswap2.tmp" $pathtoswap1 + + log_must zpool import -d $DEVICE_DIR $TESTPOOL1 + log_must check_pool_healthy $TESTPOOL1 + + # Cleanup + log_must zpool destroy $TESTPOOL1 + + log_note "" +} + +test_new_paths "$VDEV0 $VDEV1" "$VDEV0 $VDEV1" +test_new_paths "mirror $VDEV0 $VDEV1" "$VDEV0 $VDEV1" +test_new_paths "$VDEV0 log $VDEV1" "$VDEV1" +test_new_paths "raidz $VDEV0 $VDEV1 $VDEV2" "$VDEV1" + +test_swap_paths "$VDEV0 $VDEV1" "$VDEV0" "$VDEV1" +test_swap_paths "raidz $VDEV0 $VDEV1 $VDEV2" "$VDEV0" "$VDEV1" +test_swap_paths "mirror $VDEV0 $VDEV1 mirror $VDEV2 $VDEV3" \ + "$VDEV0" "$VDEV2" + +log_pass "zpool import succeeded after changing device paths." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_config_changed.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_config_changed.ksh new file mode 100755 index 000000000..92d814015 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_config_changed.ksh @@ -0,0 +1,239 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# It should be possible to rewind a pool beyond a configuration change. +# +# STRATEGY: +# 1. Create a pool. +# 2. Generate files and remember their md5sum. +# 3. Note last synced txg. +# 4. Take a snapshot to make sure old blocks are not overwritten. +# 5. Perform zpool add/attach/detach/remove operation. +# 6. Change device paths if requested and re-import pool. +# 7. Overwrite the files. +# 8. Export the pool. +# 9. Verify that we can rewind the pool to the noted txg. +# 10. Verify that the files are readable and retain their old data. +# +# DISCLAIMER: +# This test can fail since nothing guarantees that old MOS blocks aren't +# overwritten. Snapshots protect datasets and data files but not the MOS. +# sync_some_data_a_few_times interleaves file data and MOS data for a few +# txgs, thus increasing the odds that some txgs will have their MOS data +# left untouched. +# + +verify_runnable "global" + +function custom_cleanup +{ + set_vdev_validate_skip 0 + cleanup +} + +log_onexit custom_cleanup + +function test_common +{ + typeset poolcreate="$1" + typeset addvdevs="$2" + typeset attachargs="${3:-}" + typeset detachvdev="${4:-}" + typeset removevdev="${5:-}" + typeset finalpool="${6:-}" + + typeset poolcheck="$poolcreate" + + log_must zpool create $TESTPOOL1 $poolcreate + + log_must generate_data $TESTPOOL1 $MD5FILE + + # syncing a few times while writing new data increases the odds that MOS + # metadata for some of the txgs will survive + log_must sync_some_data_a_few_times $TESTPOOL1 + typeset txg + txg=$(get_last_txg_synced $TESTPOOL1) + log_must zfs snapshot -r $TESTPOOL1@snap1 + + # + # Perform config change operations + # + if [[ -n $addvdev ]]; then + log_must zpool add -f $TESTPOOL1 $addvdev + fi + if [[ -n $attachargs ]]; then + log_must zpool attach $TESTPOOL1 $attachargs + fi + if [[ -n $detachvdev ]]; then + log_must zpool detach $TESTPOOL1 $detachvdev + fi + if [[ -n $removevdev ]]; then + [[ -z $finalpool ]] && + log_fail "Must provide final pool status!" + log_must zpool remove $TESTPOOL1 $removevdev + log_must wait_for_pool_config $TESTPOOL1 "$finalpool" + fi + if [[ -n $pathstochange ]]; then + # + # Change device paths and re-import pool to update labels + # + zpool export $TESTPOOL1 + for dev in $pathstochange; do + log_must mv $dev "${dev}_new" + poolcheck=$(echo "$poolcheck" | \ + sed "s:$dev:${dev}_new:g") + done + zpool import -d $DEVICE_DIR $TESTPOOL1 + fi + + log_must overwrite_data $TESTPOOL1 "" + + log_must zpool export $TESTPOOL1 + + log_must zpool import -d $DEVICE_DIR -T $txg $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcheck" + + log_must verify_data_md5sums $MD5FILE + + # Cleanup + log_must zpool destroy $TESTPOOL1 + if [[ -n $pathstochange ]]; then + for dev in $pathstochange; do + log_must mv "${dev}_new" $dev + done + fi + # Fast way to clear vdev labels + log_must zpool create -f $TESTPOOL2 $VDEV0 $VDEV1 $VDEV2 $VDEV3 $VDEV4 + log_must zpool destroy $TESTPOOL2 + + log_note "" +} + +function test_add_vdevs +{ + typeset poolcreate="$1" + typeset addvdevs="$2" + + log_note "$0: pool '$poolcreate', add $addvdevs." + + test_common "$poolcreate" "$addvdevs" +} + +function test_attach_vdev +{ + typeset poolcreate="$1" + typeset attachto="$2" + typeset attachvdev="$3" + + log_note "$0: pool '$poolcreate', attach $attachvdev to $attachto." + + test_common "$poolcreate" "" "$attachto $attachvdev" +} + +function test_detach_vdev +{ + typeset poolcreate="$1" + typeset detachvdev="$2" + + log_note "$0: pool '$poolcreate', detach $detachvdev." + + test_common "$poolcreate" "" "" "$detachvdev" +} + +function test_attach_detach_vdev +{ + typeset poolcreate="$1" + typeset attachto="$2" + typeset attachvdev="$3" + typeset detachvdev="$4" + + log_note "$0: pool '$poolcreate', attach $attachvdev to $attachto," \ + "then detach $detachvdev." + + test_common "$poolcreate" "" "$attachto $attachvdev" "$detachvdev" +} + +function test_remove_vdev +{ + typeset poolcreate="$1" + typeset removevdev="$2" + typeset finalpool="$3" + + log_note "$0: pool '$poolcreate', remove $removevdev." + + test_common "$poolcreate" "" "" "" "$removevdev" "$finalpool" +} + +# Record txg history +is_linux && log_must set_tunable32 zfs_txg_history 100 + +# Make the devices bigger to reduce chances of overwriting MOS metadata. +increase_device_sizes $(( FILE_SIZE * 4 )) + +# Part of the rewind test is to see how it reacts to path changes +typeset pathstochange="$VDEV0 $VDEV1 $VDEV2 $VDEV3" + +log_note " == test rewind after device addition == " + +test_add_vdevs "$VDEV0" "$VDEV1" +test_add_vdevs "$VDEV0 $VDEV1" "$VDEV2" +test_add_vdevs "$VDEV0" "$VDEV1 $VDEV2" +test_add_vdevs "mirror $VDEV0 $VDEV1" "mirror $VDEV2 $VDEV3" +test_add_vdevs "$VDEV0" "raidz $VDEV1 $VDEV2 $VDEV3" +test_add_vdevs "$VDEV0" "log $VDEV1" +test_add_vdevs "$VDEV0 log $VDEV1" "$VDEV2" + +log_note " == test rewind after device attach == " + +test_attach_vdev "$VDEV0" "$VDEV0" "$VDEV1" +test_attach_vdev "mirror $VDEV0 $VDEV1" "$VDEV0" "$VDEV2" +test_attach_vdev "$VDEV0 $VDEV1" "$VDEV0" "$VDEV2" + +log_note " == test rewind after device removal == " + +# Once we remove a device it will be overlooked in the device scan, so we must +# preserve its original path +pathstochange="$VDEV0 $VDEV2" +test_remove_vdev "$VDEV0 $VDEV1 $VDEV2" "$VDEV1" "$VDEV0 $VDEV2" + +# +# Path change and detach are incompatible. Detach changes the guid of the vdev +# so we have no direct way to link the new path to an existing vdev. +# +pathstochange="" + +log_note " == test rewind after device detach == " + +test_detach_vdev "mirror $VDEV0 $VDEV1" "$VDEV1" +test_detach_vdev "mirror $VDEV0 $VDEV1 mirror $VDEV2 $VDEV3" "$VDEV1" +test_detach_vdev "$VDEV0 log mirror $VDEV1 $VDEV2" "$VDEV2" + +log_note " == test rewind after device attach followed by device detach == " + +# +# We need to disable vdev validation since once we detach VDEV1, VDEV0 will +# inherit the mirror tvd's guid and lose its original guid. +# +set_vdev_validate_skip 1 +test_attach_detach_vdev "$VDEV0" "$VDEV0" "$VDEV1" "$VDEV1" +set_vdev_validate_skip 0 + +log_pass "zpool import rewind after configuration change passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_device_replaced.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_device_replaced.ksh new file mode 100755 index 000000000..5ff1c47f3 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/import_rewind_device_replaced.ksh @@ -0,0 +1,186 @@ +#!/usr/bin/ksh -p + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.kshlib + +# +# DESCRIPTION: +# It should be possible to rewind a pool beyond a device replacement. +# +# STRATEGY: +# 1. Create a pool. +# 2. Generate files and remember their md5sum. +# 3. Sync a few times and note last synced txg. +# 4. Take a snapshot to make sure old blocks are not overwritten. +# 5. Initiate device replacement and export the pool. Special care must +# be taken so that resilvering doesn't complete before the export. +# 6. Test 1: Rewind pool to noted txg and then verify data checksums. +# Import it read-only so that we do not overwrite blocks in later txgs. +# 7. Re-import pool at latest txg and let the replacement finish. +# 8. Export the pool an remove the new device - we shouldn't need it. +# 9. Test 2: Rewind pool to noted txg and then verify data checksums. +# +# STRATEGY TO SLOW DOWN RESILVERING: +# 1. Reduce zfs_txg_timeout, which controls how long can we resilver for +# each sync. +# 2. Add data to pool +# 3. Re-import the pool so that data isn't cached +# 4. Use zinject to slow down device I/O +# 5. Trigger the resilvering +# 6. Use spa freeze to stop writing to the pool. +# 7. Clear zinject events (needed to export the pool) +# 8. Export the pool +# +# DISCLAIMER: +# This test can fail since nothing guarantees that old MOS blocks aren't +# overwritten. Snapshots protect datasets and data files but not the MOS. +# sync_some_data_a_few_times interleaves file data and MOS data for a few +# txgs, thus increasing the odds that some txgs will have their MOS data +# left untouched. +# + +verify_runnable "global" + +ZFS_TXG_TIMEOUT="" + +function custom_cleanup +{ + # Revert zfs_txg_timeout to defaults + [[ -n ZFS_TXG_TIMEOUT ]] && + log_must set_zfs_txg_timeout $ZFS_TXG_TIMEOUT + log_must rm -rf $BACKUP_DEVICE_DIR + zinject -c all + cleanup +} + +log_onexit custom_cleanup + +function test_replace_vdev +{ + typeset poolcreate="$1" + typeset replacevdev="$2" + typeset replaceby="$3" + typeset poolfinalstate="$4" + typeset zinjectdevices="$5" + typeset writedata="$6" + + log_note "$0: pool '$poolcreate', replace $replacevdev by $replaceby." + + log_must zpool create $TESTPOOL1 $poolcreate + + # generate data and checksum it + log_must generate_data $TESTPOOL1 $MD5FILE + + # add more data so that resilver takes longer + log_must write_some_data $TESTPOOL1 $writedata + + # Syncing a few times while writing new data increases the odds that + # MOS metadata for some of the txgs will survive. + log_must sync_some_data_a_few_times $TESTPOOL1 + typeset txg + txg=$(get_last_txg_synced $TESTPOOL1) + log_must zfs snapshot -r $TESTPOOL1@snap1 + + # This should not free original data. + log_must overwrite_data $TESTPOOL1 "" + + # Steps to insure resilvering happens very slowly. + log_must zpool export $TESTPOOL1 + log_must zpool import -d $DEVICE_DIR $TESTPOOL1 + typeset device + for device in $zinjectdevices ; do + log_must zinject -d $device -D 200:1 $TESTPOOL1 > /dev/null + done + log_must zpool replace $TESTPOOL1 $replacevdev $replaceby + + # We must disable zinject in order to export the pool, so we freeze + # it first to prevent writing out subsequent resilvering progress. + log_must zpool freeze $TESTPOOL1 + # Confirm pool is still replacing + log_must pool_is_replacing $TESTPOOL1 + log_must zinject -c all > /dev/null + log_must zpool export $TESTPOOL1 + + ############################################################ + # Test 1: rewind while device is resilvering. + # Import read only to avoid overwriting more recent blocks. + ############################################################ + log_must zpool import -d $DEVICE_DIR -o readonly=on -T $txg $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcreate" + + log_must verify_data_md5sums $MD5FILE + + log_must zpool export $TESTPOOL1 + + # Import pool at latest txg to finish the resilvering + log_must zpool import -d $DEVICE_DIR $TESTPOOL1 + log_must overwrite_data $TESTPOOL1 "" + log_must wait_for_pool_config $TESTPOOL1 "$poolfinalstate" + log_must zpool export $TESTPOOL1 + + # Move out the new device + log_must mv $replaceby $BACKUP_DEVICE_DIR/ + + ############################################################ + # Test 2: rewind after device has been replaced. + # Import read-write since we won't need the pool anymore. + ############################################################ + log_must zpool import -d $DEVICE_DIR -T $txg $TESTPOOL1 + log_must check_pool_config $TESTPOOL1 "$poolcreate" + + log_must verify_data_md5sums $MD5FILE + + # Cleanup + log_must zpool destroy $TESTPOOL1 + # Restore the device we moved out + log_must mv "$BACKUP_DEVICE_DIR/$(basename $replaceby)" $DEVICE_DIR/ + # Fast way to clear vdev labels + log_must zpool create -f $TESTPOOL2 $VDEV0 $VDEV1 $VDEV2 $VDEV3 $VDEV4 + log_must zpool destroy $TESTPOOL2 + + log_note "" +} + +# Record txg history +is_linux && log_must set_tunable32 zfs_txg_history 100 + +log_must mkdir -p $BACKUP_DEVICE_DIR +# Make the devices bigger to reduce chances of overwriting MOS metadata. +increase_device_sizes $(( FILE_SIZE * 4 )) + +# We set zfs_txg_timeout to 1 to reduce resilvering time at each sync. +ZFS_TXG_TIMEOUT=$(get_zfs_txg_timeout) +set_zfs_txg_timeout 1 + +test_replace_vdev "$VDEV0 $VDEV1" \ + "$VDEV1" "$VDEV2" \ + "$VDEV0 $VDEV2" \ + "$VDEV0 $VDEV1" 15 + +test_replace_vdev "mirror $VDEV0 $VDEV1" \ + "$VDEV1" "$VDEV2" \ + "mirror $VDEV0 $VDEV2" \ + "$VDEV0 $VDEV1" 10 + +test_replace_vdev "raidz $VDEV0 $VDEV1 $VDEV2" \ + "$VDEV1" "$VDEV3" \ + "raidz $VDEV0 $VDEV3 $VDEV2" \ + "$VDEV0 $VDEV1 $VDEV2" 10 + +set_zfs_txg_timeout $ZFS_TXG_TIMEOUT + +log_pass "zpool import rewind after device replacement passed." diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/setup.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_import/setup.ksh index 142771de6..d81e66636 100755 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_import/setup.ksh +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/setup.ksh @@ -69,26 +69,14 @@ log_must zfs create $TESTPOOL/$TESTFS log_must zfs set mountpoint=$TESTDIR $TESTPOOL/$TESTFS DISK2="$(echo $DISKS | nawk '{print $2}')" -if is_mpath_device $DISK2; then - echo "y" | newfs -v $DEV_DSKDIR/$DISK2 >/dev/null 2>&1 - (( $? != 0 )) && - log_untested "Unable to setup a $NEWFS_DEFAULT_FS file system" +echo "y" | newfs -v $DEV_DSKDIR/$DISK2 >/dev/null 2>&1 +(( $? != 0 )) && + log_untested "Unable to setup a $NEWFS_DEFAULT_FS file system" - [[ ! -d $DEVICE_DIR ]] && \ - log_must mkdir -p $DEVICE_DIR +[[ ! -d $DEVICE_DIR ]] && \ + log_must mkdir -p $DEVICE_DIR - log_must mount $DEV_DSKDIR/$DISK2 $DEVICE_DIR -else - log_must set_partition 0 "" $FS_SIZE $ZFS_DISK2 - echo "y" | newfs -v $DEV_DSKDIR/$ZFSSIDE_DISK2 >/dev/null 2>&1 - (( $? != 0 )) && - log_untested "Unable to setup a $NEWFS_DEFAULT_FS file system" - - [[ ! -d $DEVICE_DIR ]] && \ - log_must mkdir -p $DEVICE_DIR - - log_must mount $DEV_DSKDIR/$ZFSSIDE_DISK2 $DEVICE_DIR -fi +log_must mount $DEV_DSKDIR/$DISK2 $DEVICE_DIR i=0 while (( i < $MAX_NUM )); do diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.cfg b/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.cfg index 648f82c2b..20f43cefa 100644 --- a/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.cfg +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.cfg @@ -25,7 +25,7 @@ # # -# Copyright (c) 2012, 2015 by Delphix. All rights reserved. +# Copyright (c) 2012, 2016 by Delphix. All rights reserved. # . $STF_SUITE/include/libtest.shlib @@ -57,10 +57,8 @@ case "${#disk_array[*]}" in if ( is_mpath_device $ZFS_DISK1 ) && [[ -z $(echo $ZFS_DISK1 | awk 'substr($1,18,1)\ ~ /^[[:digit:]]+$/') ]] || ( is_real_device $ZFS_DISK1 ); then ZFSSIDE_DISK1=${ZFS_DISK1}1 - ZFSSIDE_DISK2=${ZFS_DISK2}2 elif ( is_mpath_device $ZFS_DISK1 || is_loop_device $ZFS_DISK1 ); then ZFSSIDE_DISK1=${ZFS_DISK1}p1 - ZFSSIDE_DISK2=${ZFS_DISK2}p2 else log_fail "$ZFS_DISK1 not supported for partitioning." fi @@ -71,7 +69,6 @@ case "${#disk_array[*]}" in ZFS_DISK1=${disk_array[0]} ZFSSIDE_DISK1=${ZFS_DISK1}s0 ZFS_DISK2=${disk_array[0]} - ZFSSIDE_DISK2=${ZFS_DISK2}s1 fi ;; *) @@ -96,14 +93,6 @@ case "${#disk_array[*]}" in log_fail "$ZFS_DISK1 not supported for partitioning." fi ZFS_DISK2=${disk_array[1]} - if ( is_mpath_device $ZFS_DISK2 ) && [[ -z $(echo $ZFS_DISK2 | awk 'substr($1,18,1)\ - ~ /^[[:digit:]]+$/') ]] || ( is_real_device $ZFS_DISK2 ); then - ZFSSIDE_DISK2=${ZFS_DISK2}1 - elif ( is_mpath_device $ZFS_DISK2 || is_loop_device $ZFS_DISK2 ); then - ZFSSIDE_DISK2=${ZFS_DISK2}p1 - else - log_fail "$ZFS_DISK2 not supported for partitioning." - fi else export DEV_DSKDIR="/dev" PRIMARY_SLICE=2 @@ -111,15 +100,14 @@ case "${#disk_array[*]}" in ZFS_DISK1=${disk_array[0]} ZFSSIDE_DISK1=${ZFS_DISK1}s0 ZFS_DISK2=${disk_array[1]} - ZFSSIDE_DISK2=${ZFS_DISK2}s0 fi ;; esac -export DISK_COUNT ZFS_DISK1 ZFSSIDE_DISK1 ZFS_DISK2 ZFSSIDE_DISK2 +export DISK_COUNT ZFS_DISK1 ZFSSIDE_DISK1 ZFS_DISK2 -export FS_SIZE="$((($MINVDEVSIZE / (1024 * 1024)) * 16))m" -export FILE_SIZE="$(($MINVDEVSIZE / 2))" +export FS_SIZE="$((($MINVDEVSIZE / (1024 * 1024)) * 32))m" +export FILE_SIZE="$((MINVDEVSIZE))" export SLICE_SIZE="$((($MINVDEVSIZE / (1024 * 1024)) * 2))m" export MAX_NUM=5 export GROUP_NUM=3 @@ -129,6 +117,12 @@ export DEVICE_FILE=disk export DEVICE_ARCHIVE=archive_import-test export MYTESTFILE=$STF_SUITE/include/libtest.shlib +export CPATH=/var/tmp/cachefile.$$ +export CPATHBKP=/var/tmp/cachefile.$$.bkp +export CPATHBKP2=/var/tmp/cachefile.$$.bkp2 +export MD5FILE=/var/tmp/md5sums.$$ +export MD5FILE2=/var/tmp/md5sums.$$.2 + typeset -i num=0 while (( num < $GROUP_NUM )); do DEVICE_FILES="$DEVICE_FILES ${DEVICE_DIR}/${DEVICE_FILE}$num" diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.kshlib b/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.kshlib new file mode 100644 index 000000000..bc89d8159 --- /dev/null +++ b/tests/zfs-tests/tests/functional/cli_root/zpool_import/zpool_import.kshlib @@ -0,0 +1,376 @@ +#!/usr/bin/ksh + +# +# This file and its contents are supplied under the terms of the +# Common Development and Distribution License ("CDDL"), version 1.0. +# You may only use this file in accordance with the terms of version +# 1.0 of the CDDL. +# +# A full copy of the text of the CDDL should have accompanied this +# source. A copy of the CDDL is also available via the Internet at +# http://www.illumos.org/license/CDDL. +# + +# +# Copyright (c) 2016 by Delphix. All rights reserved. +# + +. $STF_SUITE/include/libtest.shlib +. $STF_SUITE/tests/functional/cli_root/zpool_import/zpool_import.cfg + +# +# Prototype cleanup function for zpool_import tests. +# +function cleanup +{ + destroy_pool $TESTPOOL1 + + log_must rm -f $CPATH $CPATHBKP $CPATHBKP2 $MD5FILE $MD5FILE2 + + log_must rm -rf $DEVICE_DIR/* + typeset i=0 + while (( i < $MAX_NUM )); do + log_must mkfile $FILE_SIZE ${DEVICE_DIR}/${DEVICE_FILE}$i + ((i += 1)) + done + is_linux && set_tunable32 "zfs_txg_history" 0 +} + +# +# Write a bit of data and sync several times. +# This function is intended to be used by zpool rewind tests. +# +function sync_some_data_a_few_times +{ + typeset pool=$1 + typeset -i a_few_times=${2:-10} + + typeset file="/$pool/tmpfile" + for i in {0..$a_few_times}; do + dd if=/dev/urandom of=${file}_$i bs=128k count=10 + sync_pool "$pool" + done + + return 0 +} + +# +# Just write a moderate amount of data to the pool. +# +function write_some_data +{ + typeset pool=$1 + typeset files10mb=${2:-10} + + typeset ds="$pool/fillerds" + zfs create $ds + [[ $? -ne 0 ]] && return 1 + + # Create 100 MB of data + typeset file="/$ds/fillerfile" + for i in {1..$files10mb}; do + dd if=/dev/urandom of=$file.$i bs=128k count=80 + [[ $? -ne 0 ]] && return 1 + done + + return 0 +} + +# +# Create/overwrite a few datasets with files. +# Apply md5sum on all the files and store checksums in a file. +# +# newdata: overwrite existing files if false. +# md5file: file where to store md5sums +# datasetname: base name for datasets +# +function _generate_data_common +{ + typeset pool=$1 + typeset newdata=$2 + typeset md5file=$3 + typeset datasetname=$4 + + typeset -i datasets=3 + typeset -i files=5 + typeset -i blocks=10 + + [[ -n $md5file ]] && rm -f $md5file + for i in {1..$datasets}; do + ( $newdata ) && log_must zfs create "$pool/$datasetname$i" + for j in {1..$files}; do + typeset file="/$pool/$datasetname$i/file$j" + dd if=/dev/urandom of=$file bs=128k count=$blocks > /dev/null + [[ -n $md5file ]] && md5sum $file >> $md5file + done + ( $newdata ) && sync_pool "$pool" + done + + return 0 +} + +function generate_data +{ + typeset pool=$1 + typeset md5file="$2" + typeset datasetname=${3:-ds} + + _generate_data_common $pool true "$md5file" $datasetname +} + +function overwrite_data +{ + typeset pool=$1 + typeset md5file="$2" + typeset datasetname=${3:-ds} + + _generate_data_common $1 false "$md5file" $datasetname +} + +# +# Verify md5sums of every file in md5sum file $1. +# +function verify_data_md5sums +{ + typeset md5file=$1 + + if [[ ! -f $md5file ]]; then + log_note "md5 sums file '$md5file' doesn't exist" + return 1 + fi + + md5sum -c --quiet $md5file + return $? +} + +# +# Set devices size in DEVICE_DIR to $1. +# +function increase_device_sizes +{ + typeset newfilesize=$1 + + typeset -i i=0 + while (( i < $MAX_NUM )); do + log_must mkfile $newfilesize ${DEVICE_DIR}/${DEVICE_FILE}$i + ((i += 1)) + done +} + +# +# Translate vdev names returned by zpool status into more generic names. +# +# eg: mirror-2 --> mirror +# +function _translate_vdev +{ + typeset vdev=$1 + + typeset keywords="mirror replacing raidz1 raidz2 raidz3 indirect" + for word in $keywords; do + echo $vdev | egrep "^${word}-[0-9]+\$" > /dev/null + if [[ $? -eq 0 ]]; then + vdev=$word + break + fi + done + + [[ $vdev == "logs" ]] && echo "log" && return 0 + [[ $vdev == "raidz1" ]] && echo "raidz" && return 0 + + echo $vdev + return 0 +} + +# +# Check that pool configuration returned by zpool status matches expected +# configuration. Format for the check string is same as the vdev arguments for +# creating a pool +# Add -q for quiet mode. +# +# eg: check_pool_config pool1 "mirror c0t0d0s0 c0t1d0s0 log c1t1d0s0" +# +function check_pool_config +{ + typeset logfailure=true + if [[ $1 == '-q' ]]; then + logfailure=false + shift + fi + + typeset poolname=$1 + typeset expected=$2 + + typeset status + status=$(zpool status $poolname 2>&1) + if [[ $? -ne 0 ]]; then + if ( $logfailure ); then + log_note "zpool status $poolname failed: $status" + fi + return 1 + fi + + typeset actual="" + typeset began=false + printf "$status\n" | while read line; do + typeset vdev=$(echo "$line" | awk '{printf $1}') + if ( ! $began ) && [[ $vdev == NAME ]]; then + began=true + continue + fi + ( $began ) && [[ -z $vdev ]] && break; + + if ( $began ); then + [[ -z $actual ]] && actual="$vdev" && continue + vdev=$(_translate_vdev $vdev) + actual="$actual $vdev" + fi + done + + expected="$poolname $expected" + + if [[ "$actual" != "$expected" ]]; then + if ( $logfailure ); then + log_note "expected pool vdevs:" + log_note "> '$expected'" + log_note "actual pool vdevs:" + log_note "> '$actual'" + fi + return 1 + fi + + return 0 +} + +# +# Check that pool configuration returned by zpool status matches expected +# configuration within a given timeout in seconds. See check_pool_config(). +# +# eg: wait_for_pool_config pool1 "mirror c0t0d0s0 c0t1d0s0" 60 +# +function wait_for_pool_config +{ + typeset poolname=$1 + typeset expectedconfig="$2" + typeset -i timeout=${3:-60} + + timeout=$(( $timeout + $(date +%s) )) + + while (( $(date +%s) < $timeout )); do + check_pool_config -q $poolname "$expectedconfig" + [[ $? -eq 0 ]] && return 0 + sleep 3 + done + + check_pool_config $poolname "$expectedconfig" + return $? +} + +# +# Check that pool status is ONLINE +# +function check_pool_healthy +{ + typeset pool=$1 + + typeset status + status=$(zpool status $pool 2>&1) + if [[ $? -ne 0 ]]; then + log_note "zpool status $pool failed: $status" + return 1 + fi + + status=$(echo "$status" | grep "$pool" | grep -v "pool:" | \ + awk '{print $2}') + + if [[ $status != "ONLINE" ]]; then + log_note "Invalid zpool status for '$pool': '$status'" \ + "!= 'ONLINE'" + return 1 + fi + + return 0 +} + +# +# Return 0 if a device is currently being replaced in the pool. +# +function pool_is_replacing +{ + typeset pool=$1 + + zpool status $pool | grep "replacing" | grep "ONLINE" > /dev/null + + return $? +} + +function set_vdev_validate_skip +{ + set_tunable32 "vdev_validate_skip" "$1" +} + +function get_zfs_txg_timeout +{ + get_tunable "zfs_txg_timeout" +} + +function set_zfs_txg_timeout +{ + set_tunable32 "zfs_txg_timeout" "$1" +} + +function set_spa_load_verify_metadata +{ + set_tunable32 "spa_load_verify_metadata" "$1" +} + +function set_spa_load_verify_data +{ + set_tunable32 "spa_load_verify_data" "$1" +} + +function set_zfs_max_missing_tvds +{ + set_tunable32 "zfs_max_missing_tvds" "$1" +} + +# +# Use mdb to find the last txg that was synced in an active pool. +# +function get_last_txg_synced +{ + typeset pool=$1 + + if is_linux; then + txg=$(tail "/proc/spl/kstat/zfs/$pool/txgs" | + awk '$3=="C" {print $1}' | tail -1) + [[ "$txg" ]] || txg=0 + echo $txg + return 0 + fi + + typeset spas + spas=$(mdb -k -e "::spa") + [[ $? -ne 0 ]] && return 1 + + typeset spa="" + print "$spas\n" | while read line; do + typeset poolname=$(echo "$line" | awk '{print $3}') + typeset addr=$(echo "$line" | awk '{print $1}') + if [[ $poolname == $pool ]]; then + spa=$addr + break + fi + done + if [[ -z $spa ]]; then + log_fail "Couldn't find pool '$pool'" + return 1 + fi + typeset mdbcmd="$spa::print spa_t spa_ubsync.ub_txg | ::eval '.=E'" + typeset -i txg + txg=$(mdb -k -e "$mdbcmd") + [[ $? -ne 0 ]] && return 1 + + echo $txg + return 0 +} |