OpenZFS 7614, 9064 - zfs device evacuation/removal

OpenZFS 7614 - zfs device evacuation/removal
OpenZFS 9064 - remove_mirror should wait for device removal to complete

This project allows top-level vdevs to be removed from the storage pool
with "zpool remove", reducing the total amount of storage in the pool.
This operation copies all allocated regions of the device to be removed
onto other devices, recording the mapping from old to new location.
After the removal is complete, read and free operations to the removed
(now "indirect") vdev must be remapped and performed at the new location
on disk.  The indirect mapping table is kept in memory whenever the pool
is loaded, so there is minimal performance overhead when doing operations
on the indirect vdev.

The size of the in-memory mapping table will be reduced when its entries
become "obsolete" because they are no longer used by any block pointers
in the pool.  An entry becomes obsolete when all the blocks that use
it are freed.  An entry can also become obsolete when all the snapshots
that reference it are deleted, and the block pointers that reference it
have been "remapped" in all filesystems/zvols (and clones).  Whenever an
indirect block is written, all the block pointers in it will be "remapped"
to their new (concrete) locations if possible.  This process can be
accelerated by using the "zfs remap" command to proactively rewrite all
indirect blocks that reference indirect (removed) vdevs.

Note that when a device is removed, we do not verify the checksum of
the data that is copied.  This makes the process much faster, but if it
were used on redundant vdevs (i.e. mirror or raidz vdevs), it would be
possible to copy the wrong data, when we have the correct data on e.g.
the other side of the mirror.

At the moment, only mirrors and simple top-level vdevs can be removed
and no removal is allowed if any of the top-level vdevs are raidz.

Porting Notes:

* Avoid zero-sized kmem_alloc() in vdev_compact_children().

    The device evacuation code adds a dependency that
    vdev_compact_children() be able to properly empty the vdev_child
    array by setting it to NULL and zeroing vdev_children.  Under Linux,
    kmem_alloc() and related functions return a sentinel pointer rather
    than NULL for zero-sized allocations.

* Remove comment regarding "mpt" driver where zfs_remove_max_segment
  is initialized to SPA_MAXBLOCKSIZE.

  Change zfs_condense_indirect_commit_entry_delay_ticks to
  zfs_condense_indirect_commit_entry_delay_ms for consistency with
  most other tunables in which delays are specified in ms.

* ZTS changes:

    Use set_tunable rather than mdb
    Use zpool sync as appropriate
    Use sync_pool instead of sync
    Kill jobs during test_removal_with_operation to allow unmount/export
    Don't add non-disk names such as "mirror" or "raidz" to $DISKS
    Use $TEST_BASE_DIR instead of /tmp
    Increase HZ from 100 to 1000 which is more common on Linux

    removal_multiple_indirection.ksh
        Reduce iterations in order to not time out on the code
        coverage builders.

    removal_resume_export:
        Functionally, the test case is correct but there exists a race
        where the kernel thread hasn't been fully started yet and is
        not visible.  Wait for up to 1 second for the removal thread
        to be started before giving up on it.  Also, increase the
        amount of data copied in order that the removal not finish
        before the export has a chance to fail.

* MMP compatibility, the concept of concrete versus non-concrete devices
  has slightly changed the semantics of vdev_writeable().  Update
  mmp_random_leaf_impl() accordingly.

* Updated dbuf_remap() to handle the org.zfsonlinux:large_dnode pool
  feature which is not supported by OpenZFS.

* Added support for new vdev removal tracepoints.

* Test cases removal_with_zdb and removal_condense_export have been
  intentionally disabled.  When run manually they pass as intended,
  but when running in the automated test environment they produce
  unreliable results on the latest Fedora release.

  They may work better once the upstream pool import refectoring is
  merged into ZoL at which point they will be re-enabled.

Authored by: Matthew Ahrens <[email protected]>
Reviewed-by: Alex Reece <[email protected]>
Reviewed-by: George Wilson <[email protected]>
Reviewed-by: John Kennedy <[email protected]>
Reviewed-by: Prakash Surya <[email protected]>
Reviewed by: Richard Laager <[email protected]>
Reviewed by: Tim Chase <[email protected]>
Reviewed by: Brian Behlendorf <[email protected]>
Approved by: Garrett D'Amore <[email protected]>
Ported-by: Tim Chase <[email protected]>
Signed-off-by: Tim Chase <[email protected]>

OpenZFS-issue: https://www.illumos.org/issues/7614
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/f539f1eb
Closes #6900

1 files changed, 94 insertions, 1 deletions

diff --git a/include/sys/vdev_impl.h b/include/sys/vdev_impl.h
index 2ad3510ea..b933f9ab8 100644
--- a/include/sys/vdev_impl.h
+++ b/include/sys/vdev_impl.h
@@ -27,6 +27,7 @@
 #define	_SYS_VDEV_IMPL_H
 
 #include <sys/avl.h>
+#include <sys/bpobj.h>
 #include <sys/dmu.h>
 #include <sys/metaslab.h>
 #include <sys/nvpair.h>
@@ -34,6 +35,9 @@
 #include <sys/vdev.h>
 #include <sys/dkio.h>
 #include <sys/uberblock_impl.h>
+#include <sys/vdev_indirect_mapping.h>
+#include <sys/vdev_indirect_births.h>
+#include <sys/vdev_removal.h>
 #include <sys/zfs_ratelimit.h>
 
 #ifdef	__cplusplus
@@ -72,6 +76,11 @@ typedef boolean_t vdev_need_resilver_func_t(vdev_t *vd, uint64_t, size_t);
 typedef void	vdev_hold_func_t(vdev_t *vd);
 typedef void	vdev_rele_func_t(vdev_t *vd);
 
+typedef void	vdev_remap_cb_t(uint64_t inner_offset, vdev_t *vd,
+    uint64_t offset, uint64_t size, void *arg);
+typedef void	vdev_remap_func_t(vdev_t *vd, uint64_t offset, uint64_t size,
+    vdev_remap_cb_t callback, void *arg);
+
 typedef const struct vdev_ops {
 	vdev_open_func_t		*vdev_op_open;
 	vdev_close_func_t		*vdev_op_close;
@@ -82,6 +91,7 @@ typedef const struct vdev_ops {
 	vdev_need_resilver_func_t	*vdev_op_need_resilver;
 	vdev_hold_func_t		*vdev_op_hold;
 	vdev_rele_func_t		*vdev_op_rele;
+	vdev_remap_func_t		*vdev_op_remap;
 	char				vdev_op_type[16];
 	boolean_t			vdev_op_leaf;
 } vdev_ops_t;
@@ -130,6 +140,45 @@ struct vdev_queue {
 };
 
 /*
+ * On-disk indirect vdev state.
+ *
+ * An indirect vdev is described exclusively in the MOS config of a pool.
+ * The config for an indirect vdev includes several fields, which are
+ * accessed in memory by a vdev_indirect_config_t.
+ */
+typedef struct vdev_indirect_config {
+	/*
+	 * Object (in MOS) which contains the indirect mapping. This object
+	 * contains an array of vdev_indirect_mapping_entry_phys_t ordered by
+	 * vimep_src. The bonus buffer for this object is a
+	 * vdev_indirect_mapping_phys_t. This object is allocated when a vdev
+	 * removal is initiated.
+	 *
+	 * Note that this object can be empty if none of the data on the vdev
+	 * has been copied yet.
+	 */
+	uint64_t	vic_mapping_object;
+
+	/*
+	 * Object (in MOS) which contains the birth times for the mapping
+	 * entries. This object contains an array of
+	 * vdev_indirect_birth_entry_phys_t sorted by vibe_offset. The bonus
+	 * buffer for this object is a vdev_indirect_birth_phys_t. This object
+	 * is allocated when a vdev removal is initiated.
+	 *
+	 * Note that this object can be empty if none of the vdev has yet been
+	 * copied.
+	 */
+	uint64_t	vic_births_object;
+
+	/*
+	 * This is the vdev ID which was removed previous to this vdev, or
+	 * UINT64_MAX if there are no previously removed vdevs.
+	 */
+	uint64_t	vic_prev_indirect_vdev;
+} vdev_indirect_config_t;
+
+/*
  * Virtual device descriptor
  */
 struct vdev {
@@ -188,6 +237,40 @@ struct vdev {
 	uint64_t	vdev_top_zap;
 
 	/*
+	 * Values stored in the config for an indirect or removing vdev.
+	 */
+	vdev_indirect_config_t	vdev_indirect_config;
+
+	/*
+	 * The vdev_indirect_rwlock protects the vdev_indirect_mapping
+	 * pointer from changing on indirect vdevs (when it is condensed).
+	 * Note that removing (not yet indirect) vdevs have different
+	 * access patterns (the mapping is not accessed from open context,
+	 * e.g. from zio_read) and locking strategy (e.g. svr_lock).
+	 */
+	krwlock_t vdev_indirect_rwlock;
+	vdev_indirect_mapping_t *vdev_indirect_mapping;
+	vdev_indirect_births_t *vdev_indirect_births;
+
+	/*
+	 * In memory data structures used to manage the obsolete sm, for
+	 * indirect or removing vdevs.
+	 *
+	 * The vdev_obsolete_segments is the in-core record of the segments
+	 * that are no longer referenced anywhere in the pool (due to
+	 * being freed or remapped and not referenced by any snapshots).
+	 * During a sync, segments are added to vdev_obsolete_segments
+	 * via vdev_indirect_mark_obsolete(); at the end of each sync
+	 * pass, this is appended to vdev_obsolete_sm via
+	 * vdev_indirect_sync_obsolete().  The vdev_obsolete_lock
+	 * protects against concurrent modifications of vdev_obsolete_segments
+	 * from multiple zio threads.
+	 */
+	kmutex_t	vdev_obsolete_lock;
+	range_tree_t	*vdev_obsolete_segments;
+	space_map_t	*vdev_obsolete_sm;
+
+	/*
 	 * The queue depth parameters determine how many async writes are
 	 * still pending (i.e. allocated by net yet issued to disk) per
 	 * top-level (vdev_async_write_queue_depth) and the maximum allowed
@@ -356,7 +439,7 @@ extern void vdev_remove_parent(vdev_t *cvd);
  */
 extern void vdev_load_log_state(vdev_t *nvd, vdev_t *ovd);
 extern boolean_t vdev_log_state_valid(vdev_t *vd);
-extern void vdev_load(vdev_t *vd);
+extern int vdev_load(vdev_t *vd);
 extern int vdev_dtl_load(vdev_t *vd);
 extern void vdev_sync(vdev_t *vd, uint64_t txg);
 extern void vdev_sync_done(vdev_t *vd, uint64_t txg);
@@ -375,6 +458,7 @@ extern vdev_ops_t vdev_file_ops;
 extern vdev_ops_t vdev_missing_ops;
 extern vdev_ops_t vdev_hole_ops;
 extern vdev_ops_t vdev_spare_ops;
+extern vdev_ops_t vdev_indirect_ops;
 
 /*
  * Common size functions
@@ -389,6 +473,15 @@ extern void vdev_set_min_asize(vdev_t *vd);
 /* zdb uses this tunable, so it must be declared here to make lint happy. */
 extern int zfs_vdev_cache_size;
 
+/*
+ * Functions from vdev_indirect.c
+ */
+extern void vdev_indirect_sync_obsolete(vdev_t *vd, dmu_tx_t *tx);
+extern boolean_t vdev_indirect_should_condense(vdev_t *vd);
+extern void spa_condense_indirect_start_sync(vdev_t *vd, dmu_tx_t *tx);
+extern int vdev_obsolete_sm_object(vdev_t *vd);
+extern boolean_t vdev_obsolete_counts_are_precise(vdev_t *vd);
+
 #ifdef	__cplusplus
 }
 #endif