freebsd-nq/module/zfs/spa_config.c

620 lines
17 KiB
C
Raw Normal View History

2008-11-20 20:01:55 +00:00
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2011, 2020 by Delphix. All rights reserved.
* Copyright 2017 Joyent, Inc.
2008-11-20 20:01:55 +00:00
*/
#include <sys/spa.h>
#include <sys/file.h>
#include <sys/fm/fs/zfs.h>
2008-11-20 20:01:55 +00:00
#include <sys/spa_impl.h>
#include <sys/nvpair.h>
#include <sys/uio.h>
#include <sys/fs/zfs.h>
#include <sys/vdev_impl.h>
#include <sys/zfs_ioctl.h>
#include <sys/systeminfo.h>
#include <sys/sunddi.h>
#include <sys/zfeature.h>
#include <sys/zfs_file.h>
2008-11-20 20:01:55 +00:00
#ifdef _KERNEL
2009-02-18 20:51:31 +00:00
#include <sys/zone.h>
2008-11-20 20:01:55 +00:00
#endif
/*
* Pool configuration repository.
*
* Pool configuration is stored as a packed nvlist on the filesystem. By
* default, all pools are stored in /etc/zfs/zpool.cache and loaded on boot
* (when the ZFS module is loaded). Pools can also have the 'cachefile'
* property set that allows them to be stored in an alternate location until
* the control of external software.
*
* For each cache file, we have a single nvlist which holds all the
* configuration information. When the module loads, we read this information
* from /etc/zfs/zpool.cache and populate the SPA namespace. This namespace is
* maintained independently in spa.c. Whenever the namespace is modified, or
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 <mahrens@delphix.com> Reviewed-by: Alex Reece <alex@delphix.com> Reviewed-by: George Wilson <george.wilson@delphix.com> Reviewed-by: John Kennedy <john.kennedy@delphix.com> Reviewed-by: Prakash Surya <prakash.surya@delphix.com> Reviewed by: Richard Laager <rlaager@wiktel.com> Reviewed by: Tim Chase <tim@chase2k.com> Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov> Approved by: Garrett D'Amore <garrett@damore.org> Ported-by: Tim Chase <tim@chase2k.com> Signed-off-by: Tim Chase <tim@chase2k.com> OpenZFS-issue: https://www.illumos.org/issues/7614 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/f539f1eb Closes #6900
2016-09-22 16:30:13 +00:00
* the configuration of a pool is changed, we call spa_write_cachefile(), which
2008-11-20 20:01:55 +00:00
* walks through all the active pools and writes the configuration to disk.
*/
static uint64_t spa_config_generation = 1;
/*
* This can be overridden in userland to preserve an alternate namespace for
* userland pools when doing testing.
*/
char *spa_config_path = ZPOOL_CACHE;
int zfs_autoimport_disable = 1;
2008-11-20 20:01:55 +00:00
/*
* Called when the module is first loaded, this routine loads the configuration
* file into the SPA namespace. It does not actually open or load the pools; it
* only populates the namespace.
*/
void
spa_config_load(void)
{
void *buf = NULL;
nvlist_t *nvlist, *child;
nvpair_t *nvpair;
char *pathname;
zfs_file_t *fp;
zfs_file_attr_t zfa;
2008-11-20 20:01:55 +00:00
uint64_t fsize;
int err;
2008-11-20 20:01:55 +00:00
#ifdef _KERNEL
if (zfs_autoimport_disable)
return;
#endif
2008-11-20 20:01:55 +00:00
/*
* Open the configuration file.
*/
pathname = kmem_alloc(MAXPATHLEN, KM_SLEEP);
(void) snprintf(pathname, MAXPATHLEN, "%s", spa_config_path);
2008-11-20 20:01:55 +00:00
err = zfs_file_open(pathname, O_RDONLY, 0, &fp);
#ifdef __FreeBSD__
if (err)
err = zfs_file_open(ZPOOL_CACHE_BOOT, O_RDONLY, 0, &fp);
#endif
kmem_free(pathname, MAXPATHLEN);
if (err)
2008-11-20 20:01:55 +00:00
return;
if (zfs_file_getattr(fp, &zfa))
2008-11-20 20:01:55 +00:00
goto out;
fsize = zfa.zfa_size;
buf = kmem_alloc(fsize, KM_SLEEP);
2008-11-20 20:01:55 +00:00
/*
* Read the nvlist from the file.
*/
if (zfs_file_read(fp, buf, fsize, NULL) < 0)
2008-11-20 20:01:55 +00:00
goto out;
/*
* Unpack the nvlist.
*/
if (nvlist_unpack(buf, fsize, &nvlist, KM_SLEEP) != 0)
2008-11-20 20:01:55 +00:00
goto out;
/*
* Iterate over all elements in the nvlist, creating a new spa_t for
* each one with the specified configuration.
*/
mutex_enter(&spa_namespace_lock);
nvpair = NULL;
while ((nvpair = nvlist_next_nvpair(nvlist, nvpair)) != NULL) {
if (nvpair_type(nvpair) != DATA_TYPE_NVLIST)
continue;
child = fnvpair_value_nvlist(nvpair);
2008-11-20 20:01:55 +00:00
if (spa_lookup(nvpair_name(nvpair)) != NULL)
continue;
(void) spa_add(nvpair_name(nvpair), child, NULL);
2008-11-20 20:01:55 +00:00
}
mutex_exit(&spa_namespace_lock);
nvlist_free(nvlist);
out:
if (buf != NULL)
kmem_free(buf, fsize);
zfs_file_close(fp);
2008-11-20 20:01:55 +00:00
}
static int
spa_config_remove(spa_config_dirent_t *dp)
{
int error = 0;
/*
* Remove the cache file. If zfs_file_unlink() in not supported by the
* platform fallback to truncating the file which is functionally
* equivalent.
*/
error = zfs_file_unlink(dp->scd_path);
if (error == EOPNOTSUPP) {
int flags = O_RDWR | O_TRUNC;
zfs_file_t *fp;
error = zfs_file_open(dp->scd_path, flags, 0644, &fp);
if (error == 0) {
(void) zfs_file_fsync(fp, O_SYNC);
(void) zfs_file_close(fp);
}
}
return (error);
}
static int
spa_config_write(spa_config_dirent_t *dp, nvlist_t *nvl)
2008-11-20 20:01:55 +00:00
{
size_t buflen;
char *buf;
int oflags = O_RDWR | O_TRUNC | O_CREAT | O_LARGEFILE;
char *temp;
int err;
zfs_file_t *fp;
/*
* If the nvlist is empty (NULL), then remove the old cachefile.
*/
if (nvl == NULL) {
err = spa_config_remove(dp);
if (err == ENOENT)
err = 0;
return (err);
}
2008-11-20 20:01:55 +00:00
/*
* Pack the configuration into a buffer.
*/
buf = fnvlist_pack(nvl, &buflen);
temp = kmem_zalloc(MAXPATHLEN, KM_SLEEP);
2008-11-20 20:01:55 +00:00
Linux 4.2 compat: vfs_rename() The spa_config_write() function relies on the classic method of making sure updates to the /etc/zfs/zpool.cache file are atomic. It writes out a temporary version of the file and then uses vn_rename() to switch it in to place. This way there can never exist a partial version of the file, it's all or nothing. Conceptually this is a good strategy and it makes good sense for platforms where it's easy to do a rename within the kernel. Unfortunately, Linux is not one of those platforms. Even doing basic I/O to a file system from within the kernel is strongly discouraged. In order to support this at all the vn_rename() implementation ends up being complex and fragile. So fragile that recent Linux 4.2 changes have broken it. While it is possible to update vn_rename() to work with the latest kernels a better long term strategy is to stop using vn_rename() entirely. Then all this complex, fragile code can be removed. Achieving this is straight forward because config_write() is the only consumer of vn_rename(). This patch reworks spa_config_write() to update the cache file in place. The file will be truncated, written out, and then synced to disk. If an error is encountered the file will be unlinked leaving the system in a consistent state. This does expose a tiny tiny tiny window where a system could crash at exactly the wrong moment could leave a partially written cache file. However, this is highly unlikely because the cache file is 1) infrequently updated, 2) only a few kilobytes in size, and 3) written with a single vn_rdwr() call. If this were to somehow happen it poses no risk to pool. Simply removing the cache file will allow the pool to be imported cleanly. Going forward this will be even less of an issue as we intend to disable the use of a cache file by default. Bottom line not using vn_rename() allows us to make ZoL more robust against upstream kernel changes. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3653
2015-07-28 23:45:17 +00:00
/*
* Write the configuration to disk. Due to the complexity involved
* in performing a rename and remove from within the kernel the file
* is instead truncated and overwritten in place. This way we always
* have a consistent view of the data or a zero length file.
Linux 4.2 compat: vfs_rename() The spa_config_write() function relies on the classic method of making sure updates to the /etc/zfs/zpool.cache file are atomic. It writes out a temporary version of the file and then uses vn_rename() to switch it in to place. This way there can never exist a partial version of the file, it's all or nothing. Conceptually this is a good strategy and it makes good sense for platforms where it's easy to do a rename within the kernel. Unfortunately, Linux is not one of those platforms. Even doing basic I/O to a file system from within the kernel is strongly discouraged. In order to support this at all the vn_rename() implementation ends up being complex and fragile. So fragile that recent Linux 4.2 changes have broken it. While it is possible to update vn_rename() to work with the latest kernels a better long term strategy is to stop using vn_rename() entirely. Then all this complex, fragile code can be removed. Achieving this is straight forward because config_write() is the only consumer of vn_rename(). This patch reworks spa_config_write() to update the cache file in place. The file will be truncated, written out, and then synced to disk. If an error is encountered the file will be unlinked leaving the system in a consistent state. This does expose a tiny tiny tiny window where a system could crash at exactly the wrong moment could leave a partially written cache file. However, this is highly unlikely because the cache file is 1) infrequently updated, 2) only a few kilobytes in size, and 3) written with a single vn_rdwr() call. If this were to somehow happen it poses no risk to pool. Simply removing the cache file will allow the pool to be imported cleanly. Going forward this will be even less of an issue as we intend to disable the use of a cache file by default. Bottom line not using vn_rename() allows us to make ZoL more robust against upstream kernel changes. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3653
2015-07-28 23:45:17 +00:00
*/
err = zfs_file_open(dp->scd_path, oflags, 0644, &fp);
if (err == 0) {
err = zfs_file_write(fp, buf, buflen, NULL);
if (err == 0)
err = zfs_file_fsync(fp, O_SYNC);
Linux 4.2 compat: vfs_rename() The spa_config_write() function relies on the classic method of making sure updates to the /etc/zfs/zpool.cache file are atomic. It writes out a temporary version of the file and then uses vn_rename() to switch it in to place. This way there can never exist a partial version of the file, it's all or nothing. Conceptually this is a good strategy and it makes good sense for platforms where it's easy to do a rename within the kernel. Unfortunately, Linux is not one of those platforms. Even doing basic I/O to a file system from within the kernel is strongly discouraged. In order to support this at all the vn_rename() implementation ends up being complex and fragile. So fragile that recent Linux 4.2 changes have broken it. While it is possible to update vn_rename() to work with the latest kernels a better long term strategy is to stop using vn_rename() entirely. Then all this complex, fragile code can be removed. Achieving this is straight forward because config_write() is the only consumer of vn_rename(). This patch reworks spa_config_write() to update the cache file in place. The file will be truncated, written out, and then synced to disk. If an error is encountered the file will be unlinked leaving the system in a consistent state. This does expose a tiny tiny tiny window where a system could crash at exactly the wrong moment could leave a partially written cache file. However, this is highly unlikely because the cache file is 1) infrequently updated, 2) only a few kilobytes in size, and 3) written with a single vn_rdwr() call. If this were to somehow happen it poses no risk to pool. Simply removing the cache file will allow the pool to be imported cleanly. Going forward this will be even less of an issue as we intend to disable the use of a cache file by default. Bottom line not using vn_rename() allows us to make ZoL more robust against upstream kernel changes. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3653
2015-07-28 23:45:17 +00:00
zfs_file_close(fp);
if (err)
(void) spa_config_remove(dp);
Linux 4.2 compat: vfs_rename() The spa_config_write() function relies on the classic method of making sure updates to the /etc/zfs/zpool.cache file are atomic. It writes out a temporary version of the file and then uses vn_rename() to switch it in to place. This way there can never exist a partial version of the file, it's all or nothing. Conceptually this is a good strategy and it makes good sense for platforms where it's easy to do a rename within the kernel. Unfortunately, Linux is not one of those platforms. Even doing basic I/O to a file system from within the kernel is strongly discouraged. In order to support this at all the vn_rename() implementation ends up being complex and fragile. So fragile that recent Linux 4.2 changes have broken it. While it is possible to update vn_rename() to work with the latest kernels a better long term strategy is to stop using vn_rename() entirely. Then all this complex, fragile code can be removed. Achieving this is straight forward because config_write() is the only consumer of vn_rename(). This patch reworks spa_config_write() to update the cache file in place. The file will be truncated, written out, and then synced to disk. If an error is encountered the file will be unlinked leaving the system in a consistent state. This does expose a tiny tiny tiny window where a system could crash at exactly the wrong moment could leave a partially written cache file. However, this is highly unlikely because the cache file is 1) infrequently updated, 2) only a few kilobytes in size, and 3) written with a single vn_rdwr() call. If this were to somehow happen it poses no risk to pool. Simply removing the cache file will allow the pool to be imported cleanly. Going forward this will be even less of an issue as we intend to disable the use of a cache file by default. Bottom line not using vn_rename() allows us to make ZoL more robust against upstream kernel changes. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #3653
2015-07-28 23:45:17 +00:00
}
fnvlist_pack_free(buf, buflen);
kmem_free(temp, MAXPATHLEN);
return (err);
2008-11-20 20:01:55 +00:00
}
/*
* Synchronize pool configuration to disk. This must be called with the
Illumos #3956, #3957, #3958, #3959, #3960, #3961, #3962 3956 ::vdev -r should work with pipelines 3957 ztest should update the cachefile before killing itself 3958 multiple scans can lead to partial resilvering 3959 ddt entries are not always resilvered 3960 dsl_scan can skip over dedup-ed blocks if physical birth != logical birth 3961 freed gang blocks are not resilvered and can cause pool to suspend 3962 ztest should print out zfs debug buffer before exiting Reviewed by: Matthew Ahrens <mahrens@delphix.com> Reviewed by: Adam Leventhal <ahl@delphix.com> Approved by: Richard Lowe <richlowe@richlowe.net> References: https://www.illumos.org/issues/3956 https://www.illumos.org/issues/3957 https://www.illumos.org/issues/3958 https://www.illumos.org/issues/3959 https://www.illumos.org/issues/3960 https://www.illumos.org/issues/3961 https://www.illumos.org/issues/3962 illumos/illumos-gate@b4952e17e8858d3225793b28788278de9fe6038d Ported-by: Richard Yao <ryao@gentoo.org> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Porting notes: 1. zfs_dbgmsg_print() is only used in userland. Since we do not have mdb on Linux, it does not make sense to make it available in the kernel. This means that a build failure will occur if any future kernel patch depends on it. However, that is unlikely given that this functionality was added to support zdb. 2. zfs_dbgmsg_print() is only invoked for -VVV or greater log levels. This preserves the existing behavior of minimal noise when running with -V, and -VV. 3. In vdev_config_generate() the call to nvlist_alloc() was not changed to fnvlist_alloc() because we must pass KM_PUSHPAGE in the txg_sync context.
2013-08-07 20:16:22 +00:00
* namespace lock held. Synchronizing the pool cache is typically done after
* the configuration has been synced to the MOS. This exposes a window where
* the MOS config will have been updated but the cache file has not. If
* the system were to crash at that instant then the cached config may not
* contain the correct information to open the pool and an explicit import
Illumos #3956, #3957, #3958, #3959, #3960, #3961, #3962 3956 ::vdev -r should work with pipelines 3957 ztest should update the cachefile before killing itself 3958 multiple scans can lead to partial resilvering 3959 ddt entries are not always resilvered 3960 dsl_scan can skip over dedup-ed blocks if physical birth != logical birth 3961 freed gang blocks are not resilvered and can cause pool to suspend 3962 ztest should print out zfs debug buffer before exiting Reviewed by: Matthew Ahrens <mahrens@delphix.com> Reviewed by: Adam Leventhal <ahl@delphix.com> Approved by: Richard Lowe <richlowe@richlowe.net> References: https://www.illumos.org/issues/3956 https://www.illumos.org/issues/3957 https://www.illumos.org/issues/3958 https://www.illumos.org/issues/3959 https://www.illumos.org/issues/3960 https://www.illumos.org/issues/3961 https://www.illumos.org/issues/3962 illumos/illumos-gate@b4952e17e8858d3225793b28788278de9fe6038d Ported-by: Richard Yao <ryao@gentoo.org> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Porting notes: 1. zfs_dbgmsg_print() is only used in userland. Since we do not have mdb on Linux, it does not make sense to make it available in the kernel. This means that a build failure will occur if any future kernel patch depends on it. However, that is unlikely given that this functionality was added to support zdb. 2. zfs_dbgmsg_print() is only invoked for -VVV or greater log levels. This preserves the existing behavior of minimal noise when running with -V, and -VV. 3. In vdev_config_generate() the call to nvlist_alloc() was not changed to fnvlist_alloc() because we must pass KM_PUSHPAGE in the txg_sync context.
2013-08-07 20:16:22 +00:00
* would be required.
2008-11-20 20:01:55 +00:00
*/
void
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 <mahrens@delphix.com> Reviewed-by: Alex Reece <alex@delphix.com> Reviewed-by: George Wilson <george.wilson@delphix.com> Reviewed-by: John Kennedy <john.kennedy@delphix.com> Reviewed-by: Prakash Surya <prakash.surya@delphix.com> Reviewed by: Richard Laager <rlaager@wiktel.com> Reviewed by: Tim Chase <tim@chase2k.com> Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov> Approved by: Garrett D'Amore <garrett@damore.org> Ported-by: Tim Chase <tim@chase2k.com> Signed-off-by: Tim Chase <tim@chase2k.com> OpenZFS-issue: https://www.illumos.org/issues/7614 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/f539f1eb Closes #6900
2016-09-22 16:30:13 +00:00
spa_write_cachefile(spa_t *target, boolean_t removing, boolean_t postsysevent)
2008-11-20 20:01:55 +00:00
{
spa_config_dirent_t *dp, *tdp;
nvlist_t *nvl;
char *pool_name;
boolean_t ccw_failure;
int error = 0;
2008-11-20 20:01:55 +00:00
ASSERT(MUTEX_HELD(&spa_namespace_lock));
if (!(spa_mode_global & SPA_MODE_WRITE))
2009-01-15 21:59:39 +00:00
return;
/*
* Iterate over all cachefiles for the pool, past or present. When the
* cachefile is changed, the new one is pushed onto this list, allowing
* us to update previous cachefiles that no longer contain this pool.
*/
ccw_failure = B_FALSE;
for (dp = list_head(&target->spa_config_list); dp != NULL;
dp = list_next(&target->spa_config_list, dp)) {
spa_t *spa = NULL;
if (dp->scd_path == NULL)
continue;
/*
* Iterate over all pools, adding any matching pools to 'nvl'.
*/
nvl = NULL;
while ((spa = spa_next(spa)) != NULL) {
/*
* Skip over our own pool if we're about to remove
* ourselves from the spa namespace or any pool that
* is readonly. Since we cannot guarantee that a
* readonly pool would successfully import upon reboot,
* we don't allow them to be written to the cache file.
*/
if ((spa == target && removing) ||
!spa_writeable(spa))
continue;
mutex_enter(&spa->spa_props_lock);
tdp = list_head(&spa->spa_config_list);
if (spa->spa_config == NULL ||
tdp == NULL ||
tdp->scd_path == NULL ||
strcmp(tdp->scd_path, dp->scd_path) != 0) {
mutex_exit(&spa->spa_props_lock);
continue;
}
if (nvl == NULL)
nvl = fnvlist_alloc();
if (spa->spa_import_flags & ZFS_IMPORT_TEMP_NAME)
pool_name = fnvlist_lookup_string(
spa->spa_config, ZPOOL_CONFIG_POOL_NAME);
else
pool_name = spa_name(spa);
fnvlist_add_nvlist(nvl, pool_name, spa->spa_config);
mutex_exit(&spa->spa_props_lock);
}
error = spa_config_write(dp, nvl);
if (error != 0)
ccw_failure = B_TRUE;
nvlist_free(nvl);
}
2008-11-20 20:01:55 +00:00
if (ccw_failure) {
/*
* Keep trying so that configuration data is
* written if/when any temporary filesystem
* resource issues are resolved.
*/
if (target->spa_ccw_fail_time == 0) {
(void) zfs_ereport_post(
FM_EREPORT_ZFS_CONFIG_CACHE_WRITE,
target, NULL, NULL, NULL, 0);
}
target->spa_ccw_fail_time = gethrtime();
spa_async_request(target, SPA_ASYNC_CONFIG_UPDATE);
} else {
/*
* Do not rate limit future attempts to update
* the config cache.
*/
target->spa_ccw_fail_time = 0;
}
2008-11-20 20:01:55 +00:00
/*
* Remove any config entries older than the current one.
2008-11-20 20:01:55 +00:00
*/
dp = list_head(&target->spa_config_list);
while ((tdp = list_next(&target->spa_config_list, dp)) != NULL) {
list_remove(&target->spa_config_list, tdp);
if (tdp->scd_path != NULL)
spa_strfree(tdp->scd_path);
kmem_free(tdp, sizeof (spa_config_dirent_t));
2008-11-20 20:01:55 +00:00
}
spa_config_generation++;
if (postsysevent)
spa_event_notify(target, NULL, NULL, ESC_ZFS_CONFIG_SYNC);
2008-11-20 20:01:55 +00:00
}
/*
* Sigh. Inside a local zone, we don't have access to /etc/zfs/zpool.cache,
* and we don't want to allow the local zone to see all the pools anyway.
* So we have to invent the ZFS_IOC_CONFIG ioctl to grab the configuration
* information for all pool visible within the zone.
*/
nvlist_t *
spa_all_configs(uint64_t *generation)
{
nvlist_t *pools;
spa_t *spa = NULL;
2008-11-20 20:01:55 +00:00
if (*generation == spa_config_generation)
return (NULL);
pools = fnvlist_alloc();
2008-11-20 20:01:55 +00:00
mutex_enter(&spa_namespace_lock);
while ((spa = spa_next(spa)) != NULL) {
if (INGLOBALZONE(curproc) ||
zone_dataset_visible(spa_name(spa), NULL)) {
mutex_enter(&spa->spa_props_lock);
fnvlist_add_nvlist(pools, spa_name(spa),
spa->spa_config);
mutex_exit(&spa->spa_props_lock);
2008-11-20 20:01:55 +00:00
}
}
*generation = spa_config_generation;
mutex_exit(&spa_namespace_lock);
2008-11-20 20:01:55 +00:00
return (pools);
}
void
spa_config_set(spa_t *spa, nvlist_t *config)
{
mutex_enter(&spa->spa_props_lock);
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 <pavel.zakharov@delphix.com> Reviewed by: George Wilson <george.wilson@delphix.com> Reviewed by: Matthew Ahrens <mahrens@delphix.com> Reviewed by: Andrew Stormont <andyjstormont@gmail.com> Approved by: Hans Rosenfeld <rosenfeld@grumpf.hope-2000.org> Ported-by: Tim Chase <tim@chase2k.com> Signed-off-by: Tim Chase <tim@chase2k.com> OpenZFS-issue: https://illumos.org/issues/9075 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/619c0123 Closes #7459
2016-07-22 14:39:36 +00:00
if (spa->spa_config != NULL && spa->spa_config != config)
nvlist_free(spa->spa_config);
2008-11-20 20:01:55 +00:00
spa->spa_config = config;
mutex_exit(&spa->spa_props_lock);
2008-11-20 20:01:55 +00:00
}
/*
* Generate the pool's configuration based on the current in-core state.
*
2008-11-20 20:01:55 +00:00
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
boolean_t locked = B_FALSE;
uint64_t split_guid;
char *pool_name;
2008-11-20 20:01:55 +00:00
if (vd == NULL) {
2008-11-20 20:01:55 +00:00
vd = rvd;
locked = B_TRUE;
spa_config_enter(spa, SCL_CONFIG | SCL_STATE, FTAG, RW_READER);
}
ASSERT(spa_config_held(spa, SCL_CONFIG | SCL_STATE, RW_READER) ==
(SCL_CONFIG | SCL_STATE));
2008-11-20 20:01:55 +00:00
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
/*
* Originally, users had to handle spa namespace collisions by either
* exporting the already imported pool or by specifying a new name for
* the pool with a conflicting name. In the case of root pools from
* virtual guests, neither approach to collision resolution is
* reasonable. This is addressed by extending the new name syntax with
* an option to specify that the new name is temporary. When specified,
* ZFS_IMPORT_TEMP_NAME will be set in spa->spa_import_flags to tell us
* to use the previous name, which we do below.
*/
if (spa->spa_import_flags & ZFS_IMPORT_TEMP_NAME) {
VERIFY0(nvlist_lookup_string(spa->spa_config,
ZPOOL_CONFIG_POOL_NAME, &pool_name));
} else
pool_name = spa_name(spa);
config = fnvlist_alloc();
fnvlist_add_uint64(config, ZPOOL_CONFIG_VERSION, spa_version(spa));
fnvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME, pool_name);
fnvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE, spa_state(spa));
fnvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG, txg);
fnvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID, spa_guid(spa));
fnvlist_add_uint64(config, ZPOOL_CONFIG_ERRATA, spa->spa_errata);
if (spa->spa_comment != NULL)
fnvlist_add_string(config, ZPOOL_CONFIG_COMMENT,
spa->spa_comment);
hostid = spa_get_hostid(spa);
if (hostid != 0)
fnvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID, hostid);
fnvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME, utsname()->nodename);
2008-11-20 20:01:55 +00:00
int config_gen_flags = 0;
2008-11-20 20:01:55 +00:00
if (vd != rvd) {
fnvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid);
fnvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid);
2008-11-20 20:01:55 +00:00
if (vd->vdev_isspare)
fnvlist_add_uint64(config,
ZPOOL_CONFIG_IS_SPARE, 1ULL);
2008-11-20 20:01:55 +00:00
if (vd->vdev_islog)
fnvlist_add_uint64(config,
ZPOOL_CONFIG_IS_LOG, 1ULL);
2008-11-20 20:01:55 +00:00
vd = vd->vdev_top; /* label contains top config */
} else {
/*
* Only add the (potentially large) split information
* in the mos config, and not in the vdev labels
*/
if (spa->spa_config_splitting != NULL)
fnvlist_add_nvlist(config, ZPOOL_CONFIG_SPLIT,
spa->spa_config_splitting);
fnvlist_add_boolean(config, ZPOOL_CONFIG_HAS_PER_VDEV_ZAPS);
config_gen_flags |= VDEV_CONFIG_MOS;
2008-11-20 20:01:55 +00:00
}
/*
* Add the top-level config. We even add this on pools which
* don't support holes in the namespace.
*/
vdev_top_config_generate(spa, config);
/*
* If we're splitting, record the original pool's guid.
*/
if (spa->spa_config_splitting != NULL &&
nvlist_lookup_uint64(spa->spa_config_splitting,
ZPOOL_CONFIG_SPLIT_GUID, &split_guid) == 0) {
fnvlist_add_uint64(config, ZPOOL_CONFIG_SPLIT_GUID, split_guid);
}
nvroot = vdev_config_generate(spa, vd, getstats, config_gen_flags);
fnvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot);
2008-11-20 20:01:55 +00:00
nvlist_free(nvroot);
/*
* Store what's necessary for reading the MOS in the label.
*/
fnvlist_add_nvlist(config, ZPOOL_CONFIG_FEATURES_FOR_READ,
spa->spa_label_features);
if (getstats && spa_load_state(spa) == SPA_LOAD_NONE) {
ddt_histogram_t *ddh;
ddt_stat_t *dds;
ddt_object_t *ddo;
ddh = kmem_zalloc(sizeof (ddt_histogram_t), KM_SLEEP);
ddt_get_dedup_histogram(spa, ddh);
fnvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_HISTOGRAM,
(uint64_t *)ddh, sizeof (*ddh) / sizeof (uint64_t));
kmem_free(ddh, sizeof (ddt_histogram_t));
ddo = kmem_zalloc(sizeof (ddt_object_t), KM_SLEEP);
ddt_get_dedup_object_stats(spa, ddo);
fnvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_OBJ_STATS,
(uint64_t *)ddo, sizeof (*ddo) / sizeof (uint64_t));
kmem_free(ddo, sizeof (ddt_object_t));
dds = kmem_zalloc(sizeof (ddt_stat_t), KM_SLEEP);
ddt_get_dedup_stats(spa, dds);
fnvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_STATS,
(uint64_t *)dds, sizeof (*dds) / sizeof (uint64_t));
kmem_free(dds, sizeof (ddt_stat_t));
}
if (locked)
spa_config_exit(spa, SCL_CONFIG | SCL_STATE, FTAG);
2008-11-20 20:01:55 +00:00
return (config);
}
/*
* Update all disk labels, generate a fresh config based on the current
* in-core state, and sync the global config cache (do not sync the config
* cache if this is a booting rootpool).
*/
void
2009-08-18 18:43:27 +00:00
spa_config_update(spa_t *spa, int what)
2008-11-20 20:01:55 +00:00
{
vdev_t *rvd = spa->spa_root_vdev;
uint64_t txg;
int c;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER);
2008-11-20 20:01:55 +00:00
txg = spa_last_synced_txg(spa) + 1;
if (what == SPA_CONFIG_UPDATE_POOL) {
vdev_config_dirty(rvd);
} else {
/*
* If we have top-level vdevs that were added but have
* not yet been prepared for allocation, do that now.
* (It's safe now because the config cache is up to date,
* so it will be able to translate the new DVAs.)
* See comments in spa_vdev_add() for full details.
*/
for (c = 0; c < rvd->vdev_children; c++) {
vdev_t *tvd = rvd->vdev_child[c];
/*
* Explicitly skip vdevs that are indirect or
* log vdevs that are being removed. The reason
* is that both of those can have vdev_ms_array
* set to 0 and we wouldn't want to change their
* metaslab size nor call vdev_expand() on them.
*/
if (!vdev_is_concrete(tvd) ||
(tvd->vdev_islog && tvd->vdev_removing))
continue;
if (tvd->vdev_ms_array == 0)
2009-07-02 22:44:48 +00:00
vdev_metaslab_set_size(tvd);
vdev_expand(tvd, txg);
2008-11-20 20:01:55 +00:00
}
}
spa_config_exit(spa, SCL_ALL, FTAG);
2008-11-20 20:01:55 +00:00
/*
* Wait for the mosconfig to be regenerated and synced.
*/
txg_wait_synced(spa->spa_dsl_pool, txg);
/*
* Update the global config cache to reflect the new mosconfig.
*/
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 <mahrens@delphix.com> Reviewed-by: Alex Reece <alex@delphix.com> Reviewed-by: George Wilson <george.wilson@delphix.com> Reviewed-by: John Kennedy <john.kennedy@delphix.com> Reviewed-by: Prakash Surya <prakash.surya@delphix.com> Reviewed by: Richard Laager <rlaager@wiktel.com> Reviewed by: Tim Chase <tim@chase2k.com> Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov> Approved by: Garrett D'Amore <garrett@damore.org> Ported-by: Tim Chase <tim@chase2k.com> Signed-off-by: Tim Chase <tim@chase2k.com> OpenZFS-issue: https://www.illumos.org/issues/7614 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/f539f1eb Closes #6900
2016-09-22 16:30:13 +00:00
if (!spa->spa_is_root) {
spa_write_cachefile(spa, B_FALSE,
what != SPA_CONFIG_UPDATE_POOL);
}
2008-11-20 20:01:55 +00:00
if (what == SPA_CONFIG_UPDATE_POOL)
2009-08-18 18:43:27 +00:00
spa_config_update(spa, SPA_CONFIG_UPDATE_VDEVS);
2008-11-20 20:01:55 +00:00
}
EXPORT_SYMBOL(spa_config_load);
EXPORT_SYMBOL(spa_all_configs);
EXPORT_SYMBOL(spa_config_set);
EXPORT_SYMBOL(spa_config_generate);
EXPORT_SYMBOL(spa_config_update);
/* BEGIN CSTYLED */
#ifdef __linux__
/* string sysctls require a char array on FreeBSD */
ZFS_MODULE_PARAM(zfs_spa, spa_, config_path, STRING, ZMOD_RD,
"SPA config file (/etc/zfs/zpool.cache)");
#endif
ZFS_MODULE_PARAM(zfs, zfs_, autoimport_disable, INT, ZMOD_RW,
"Disable pool import at module load");
/* END CSTYLED */