IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
1) Since the min|maxrecoveryrate args are size_kb_ARGs and they
are recorded (and sent to the kernel) in terms of kB/sec/disk,
we must back out the factor multiple done by size_kb_arg. This
is already performed by 'lvcreate' for these arguments.
2) Allow all RAID types, not just RAID1, to change these values.
3) Add min|maxrecoveryrate_ARG to the list of 'update_partial_unsafe'
commands so that lvchange will not complain about needing at
least one of a certain set of arguments and failing.
4) Add tests that check that these values can be set via lvchange
and lvcreate and that 'lvs' reports back the proper results.
gcc -O2 v4.8 on 32 bit architecture is causing a bug in parameter
passing. It does not happen with -01 nor -O0.
The problematic part of the code was strlen use in config.c in
the config_def_check fn and the call for _config_def_check_tree in it:
<snip>
rplen = strlen(rp);
if (!_config_def_check_tree(handle, vp, vp + strlen(vp), rp, rp + rplen, CFG_PATH_MAX_LEN - rplen, cn, cmd->cft_def_hash)) ...
</snip>
If compiled with -O0 (correct):
Breakpoint 1, config_def_check (cmd=0x819b050, handle=0x81a04f8) at config/config.c:775
(gdb) p vp
$1 = 0x8189ee0 <_cfg_path> "config"
(gdb) p strlen(vp)
$2 = 6
(gdb)
_config_def_check_tree (handle=0x81a04f8, vp=0x8189ee0 <_cfg_path>
"config", pvp=0x8189ee6 <_cfg_path+6> "", rp=0xbfffe1e8 "config",
prp=0xbfffe1ee "", buf_size=58, root=0x81a2568, ht=0x81a65
48) at config/config.c:680
(gdb) p vp
$4 = 0x8189ee0 <_cfg_path> "config"
(gdb) p pvp
$5 = 0x8189ee6 <_cfg_path+6> ""
If compiled with -O2 (incorrect):
Breakpoint 1, config_def_check (cmd=cmd@entry=0x8183050, handle=0x81884f8) at config/config.c:775
(gdb) p vp
$1 = 0x8172fc0 <_cfg_path> "config"
(gdb) p strlen(vp)
$2 = 6
(gdb) p vp + strlen(vp)
$3 = 0x8172fc6 <_cfg_path+6> ""
(gdb)
_config_def_check_tree (handle=handle@entry=0x81884f8, pvp=0x8172fc7
<_cfg_path+7> "host_list", rp=rp@entry=0xbffff190 "config",
prp=prp@entry=0xbffff196 "", buf_size=buf_size@entry=58, ht=0x
818e548, root=0x818a568, vp=0x8172fc0 <_cfg_path> "config") at
config/config.c:674
(gdb) p pvp
$4 = 0x8172fc7 <_cfg_path+7> "host_list"
The difference is in passing the "pvp" arg for _config_def_check_tree.
While in the correct case, the value of _cfg_path+6 is passed
(the result of vp + strlen(vp) - see the snippet of the code above),
in the incorrect case, this value is increased by 1 to _cfg_path+7,
hence totally malforming the string that is being processed.
This ends up with incorrect validation check and incorrect warning
messages are issued like:
"Configuration setting "config/checks" has invalid type. Found integer, expected section."
To workaround this issue, remove the "static" qualifier from the
"static char _cfg_path[CFG_PATH_MAX_LEN]". This causes the optimalizer
to be less aggressive (also shuffling the arg list for
_config_def_check_tree call helps).
If there is no RAID support in the kernel but the default mirror
segtype is "raid1", converting legacy mirrors can be problematic.
For example, changing the log type or converting a mirror to a linear
LV does not require the RAID modules to be present. However, because
lp->segtype is set to be RAID1 by the configuration file, the command
fails.
We should only be setting lp->segtype when converting mirrors if it is
going to change (e.g. to linear or between mirror types).
When creating a new thin pool and there's no profile requested
via "lvcreate --profile ...", inherit any VG profile if it's attached.
Currently this applies to these settings:
allocation/thin_pool_chunk_size
allocation/thin_pool_discards
allocation/thin_pool_zero
When creating a timeout thread for snapshots, the thread is not
tracked and thus never joined. This means that the exit status
of the timeout thread is held indefinitely. Saves a bit of
memory to set PTHREAD_CREATE_DETACHED when creating this thread.
I've also added pthread_attr_init|destroy to setup the creation
pthread_attr_t.
Reported-by: NeilBrown <neilb@suse.de>
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Initial basic support for repair.
It currently takes pool metadata spare volume, which
is used for recovery. New spare is created if the volume
is successfuly repaired.
After the operation the previous _tmeta volume is moved
into _tmeta%d volume and if everything is ok, this volume
could be removed.
New _tmeta needs to be pvmoved to proper place and also
converted to i.e. mirror if it should be mirrored.
Later version will try to automate some steps here.
Add new configure lvm.conf options for binaries thin_repair
and thin_dump.
Those are part of device-mapper-persistent-data package
and will be used for recovery of thin_pool.
The PREFERRED allocation mechanism requires the number of areas in the
previous LV segment to match the number in the new segment being
allocated. If they do not match, the code may crash.
E.g. https://bugzilla.redhat.com/989347
Introduce A_AREA_COUNT_MATCHES and when not set avoid referring
to the previous segment with the contiguous and cling policies.
When using a global_filter and if this filter is incorrectly
specified, we ended up with a segfault:
raw/~ $ pvs
Invalid filter pattern "r|/dev/sda".
Segmentation fault (core dumped)
In the example above a closing '|' character is missing at the end
of the regex. The segfault itself was caused by trying to destroy
the same filter twice in _init_filters fn within the error path
(the "bad" goto target):
bad:
if (f3)
f3->destroy(f3);
if (f4)
f4->destroy(f4);
Where f3 is the composite filter (sysfs + regex + type + md + mpath filter)
and f4 is the persistent filter which encompasses this composite filter
within persistent filter's 'real' field in 'struct pfilter'.
So in the end, we need to destroy the persistent filter only as
this will also destroy any 'real' filter attached to it.
363 files changed, 19863 insertions(+), 9055 deletions(-)
This is a very large release - so expect bugs!
Please treat this release like a release candidate.
Changes to the external interfaces since 2.02.98 are not yet frozen.
Updated releases will follow quickly (days not weeks) as any problems
are handled.
The --type mirror requires -m/--mirrrors:
lvconvert --type mirror vg/lvol0
--type mirror requires -m/--mirrors
Run `lvconvert --help' for more information.
The --type raid* is allowed (the checks already existed):
lvconvert --type raid10 vg/lvol0
Converting the segment type for vg/lvol0 from linear to raid10 is not yet supported.
The --type snapshot is a synonym to -s/--snapshot:
lvconvert -s vg/lvol0 vg/lvol1
Logical volume lvol1 converted to snapshot.
lvconvert --type snapshot vg/lvol0 vg/lvol1
Logical volume lvol1 converted to snapshot.
All the other segment types are not supported, e.g.:
lvconvert --type zero vg/lvol0
Conversion using --type zero is not supported.
Run `lvconvert --help' for more information.
The status printed for dm-raid targets on older kernels does not include
the syncaction field. This is handled by dev_manager_raid_status() just
fine by populating the raid status structure with NULL for that field.
However, lv_raid_sync_action() does not properly handle that field being
NULL. So, check for it and return 0 if it is NULL.
Pool creation involves clearing of metadata device
which triggers udev watch rule we cannot udev synchronize with
in current code.
This metadata devices needs to be activated localy,
so in cluster mode deactivation and reactivation
is always needed.
However for non-clustered mode we may reload table
via suspend/resume path which avoids collision with
udev watch rule which was occasionaly triggering
retry deactivation loop.
Code has been also split into 2 separate code paths
for thin pools and thin volumes which improved readability
of the code as well.
Deactivation has been moved out of extend_pool() and
decision is now in _lv_create_an_lv() which knows
the change mode.
The new lvm2-activation-net.service activates LVM volumes
after network-attached devices are set up (iSCSI and FCoE)
if lvmetad is disabled and hence the autoactivation is not
used.
Created dlid for test is not needed afterward, so lower a memory
usage of this call is repeatedly used for building some large tree.
TODO: create function to use given buffer on stack as much cheaper.
Code needs to check if the layer origin device is suspended,
It's valid to create thinvolume snapshot of thinvolume which is also
used as an old-style snapshot. In this case we need to check -real
is suspended.
When adding origin_only - add only layer thin volume.
(in case it's also old-snapshot add only -real device)
Remove backup() call from update_pool_lv() as it's been there
duplicated and preperly order backup() call after lvresize,
so there is just one such call.
If the thin pool is known to be active, messages can be passed
to the pool even when the created thin volume is not going to be
activated.
So we do not need to stack large list of message and validate
and catch creation errors earlier in this case.
Replace the test for valid activation combination with simpler list of
deactivation combinations.
Normally, the lvm dumpconfig processes only the configuration tree
that is at the top of the cascade. Considering the cascade is:
CONFIG_STRING -> CONFIG_PROFILE -> CONFIG_MERGED_FILES/CONFIG_FILE
...then:
(dumpconfig of lvm.conf only)
raw/~ $ lvm dumpconfig allocation
allocation {
maximise_cling=1
mirror_logs_require_separate_pvs=0
thin_pool_metadata_require_separate_pvs=0
thin_pool_chunk_size=64
}
(dumpconfig of selected profile configuration only)
raw/~ $ lvm dumpconfig --profile test allocation
allocation {
thin_pool_chunk_size=8
thin_pool_discards="passdown"
thin_pool_zero=1
}
(dumpconfig of given --config configuration only)
raw/~ $ lvm dumpconfig --config 'allocation{thin_pool_chunk_size=16}' allocation
allocation {
thin_pool_chunk_size=16
}
The --mergedconfig option causes the configuration cascade to be
merged before processing it with dumpconfig:
(dumpconfig of merged selected profile and lvm.conf)
raw/~ $ lvm dumpconfig --profile test allocation --mergedconfig
allocation {
maximise_cling=1
thin_pool_zero=1
thin_pool_discards="passdown"
mirror_logs_require_separate_pvs=0
thin_pool_metadata_require_separate_pvs=0
thin_pool_chunk_size=8
}
(dumpconfig merged given --config and selected profile and lvm.conf)
raw/~ $ lvm dumpconfig --profile test --config 'allocation{thin_pool_chunk_size=16}' allocation --mergedconfig
allocation {
maximise_cling=1
thin_pool_zero=1
thin_pool_discards="passdown"
mirror_logs_require_separate_pvs=0
thin_pool_metadata_require_separate_pvs=0
thin_pool_chunk_size=16
}
Hence with the --mergedconfig, we are able to see the
configuration that is actually used when processing any
LVM command while using any combination of --config/--profile
options together with lvm.conf file.
Start separating the validation from the action in the basic lvresize
code moved to the library.
Remove incorrect use of command line error codes from lvresize library
functions. Move errors.h to tools directory to reinforce this,
exporting public versions of the error codes in lvm2cmd.h for dmeventd
plugins to use.
Document following items:
configuration cascade (man lvm.conf)
--profile ProfileName (man lvm)
--detachprofile (man vgchange/lvchange)
-o vg_profile/lv_profile (man vgs/lvs)
Also document --config a bit so we can see where it fits in the
configuration cascade - will be documented more in following commit...
Fix and improve handling on sigint.
Always check for signal presence *before* calling of command,
so it will not call the command when break was hit.
If the command has been finished succesfully there is
no problem to mark the command ok and not report interrupt at all.
Fix cuple related stack; reports and assignments.
Do not keep multiple archives for the executed command.
Reuse the ALLOCATABLE_PV from pv status for
ARCHIVED_VG vg status. Mark VG with the bit with the
first archivation.
If the user would upconvert a linear LV to a mirror without specifying
the segment type ("--type mirror" vs "--type raid1"), the "mirror"
segment type would be chosen without consulting the 'default_mirror_segtype'
setting in lvm.conf. This is now used as the basis for determining
which should be used if left unspecified.
Check for enough space in preallocated buffer.
Fixes problem, when lvm code started to suddenly allocate
too big memory chunks.
TODO: lvmetad protocol should announce needed size ahead,
so if metadata have 1MB we are not reallocating memory...
When vgname has not existed in metadata, it has crashed on double free
in format_instance destroy() - since VG was created, used FID and was
released - which also released FID, so further use was accessing bad
memory.
Fix it for this code path before release_vg() so FID will exists
when _vg_read_file_name() returns NULL.
Support vgsplit for VGs with thin pools and thin volumes.
In case the thin data and thin metadata volumes are moved to a new VG,
move there also all related thin volumes and check that external origins
are also present in this new VG.
We use mpath filtering (enabled by devices/multipath_component_detection=1
lvm.conf setting) to avoid a situation in which we could end up with
duplicate PVs found. We need to filter out the mpath components and
use only the top-level multipath mapping instead for PV scans.
However, if the there are partitions on multipath components, we need
to filter out these partitions. This patch fixes it so those
partitions found on multipath components are filtered as well.
For example, let's consider following configuration:
The sda and sdb are mpath components, sda1 and sdb1 the partitions
on these components, mpath-test the mpath mapping and mpath-test1
the partition mapping - created automatically by kpartx right
after mpath-test creation. The PV resides on top.
(LVM PV)
|
mpath-test1
|
mpath-test
|
sda1 ---------- sdb1
\ | |/
sda sdb
E.g. for sda1 and sdb1, the code will detect this and it skips
the partition that belongs to the multipath component:
<snippet from the log>
#filters/filter-mpath.c:156 /dev/sda1: Device is a partition, using primary device /dev/sda for mpath component detection
130 #ioctl/libdm-iface.c:1724 dm status (253:2) OF[16384](*1)
131 #filters/filter-mpath.c:196 /dev/sda1: Skipping mpath component device
</snippet from the log>
Othewise, we'd see the same PV label on sda1/sdb1 and mpath-test1
at the same time ending up with "Duplicate PV found...".
Add support for lvresize of thin pool metadata device.
lvresize --poolmetadatasize +20 vgname/thinpool_lv
or
lvresize -L +20 vgname/thinpool_lv_tmeta
Where the second one allows all the args for resize (striping...)
and the first option resizes accoding to the last metadata lv segment.
Giving volume type information about being 'metadata' type of volume
has higher priority then i.e. 'mirror' or 'thin' flag - for those
type we have 'target attr' (7th. field).
Last commit made dump filter only partially composable.
Add remaining functionality and also support composable wipe,
which is needed, when i.e. vgscan needs to remove cache.
(in release fix)
Add a generic dump operation to filters and make the composite filter call
through to its components. Previously, when global filter was set, the code
would treat the toplevel composite filter's private area as if it belonged a
persistent filter, trying to write nonsense into a non-sensical file.
Also deal with NULL cmd->filter gracefully.
The global filter in system's lvm.conf may conflict with the custom filter we
set up in vgimportclone (they can easily fail to intersect). Since we explicitly
avoid talking to lvmetad in vgimportclone, it is safe and reasonable to do so.
There is no point in creation of 2chunks snapshot,
since the snapshot is invalidated immeditelly with the first write
as there is no free chunk for COW blocks
(2 chunks are used by the snap header and the 1st. metadata chunk).
Enhance error message about the lowest usable size.
Instead of seeing wierd overflows inside the lvm code,
giving false error messages, kill the user experiment in the begining.
Who needs to use more then 16EiB with lvm2 and 64bit anyway...
Avoid hitting memory corruption (double free) in code path,
where PV FID has been already destroyed and the released pointer
was left in PV structure and could have been tried to be released
from there 2nd. time with final context destruction.
Check for mounted fs also for vgchange command, not just lvchange.
NOTE: Code is using lv_info() just like lvs_in_vg_opened().
It should be probably converted into lv_is_active_locally().
There are places where 'lv_is_active' was being used where it was
more correct to use 'lv_is_active_locally'. For example, when checking
for the existance of a kernel instance before asking for its status.
Most of the time these would work correctly. (RAID is only allowed on
non-clustered VGs at the moment, which means that 'lv_is_active' and
'lv_is_active_locally' would give the same result.) However, it is
more correct to use the proper variant and it helps with future
scenarios where targets might be allowed exclusively (or clustered) in
a cluster VG.
Accept --yes on all commands, even ones that don't today have prompts,
so that test scripts that don't care about interactive prompts no
longer need to deal with them.
But continue to mention --yes only in the command prototypes that
actually use it.
This fixes a long standing regression since LVM2 2.02.74 (commit 4efb1d9c,
"Update heuristic used for default and detected data alignment.")
The default PE alignment could be used (via MAX()) even if it was
determined that the device's MD stripe width, or minimal_io_size or
optimal_io_size were not factors of the default PE alignment (either 64K
or the newer default of 1MB, etc). This bug would manifest if the
default PE alignment was larger than the overriding hint that the
device provided (e.g. default of 1MB vs optimal_io_size of 768K).
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
If the dm_realloc would fail, the already allocate _maps_buffer
memory would have been lost (overwritten with NULL).
Fix this by using temporary line buffer.
Also add a minor cleanup to set end of buffer to '\0',
only when we really know the file size fits the preallocated buffer.
Setting the cmd->default_settings.udev_fallback also requires DM
driver version check. However, this caused useless mapper/control
access with ioctl if not needed actually. For example if we're not
using activation code, we don't need to know the udev_fallback as
there's no node and symlink processing.
For example, this premature mapper/control access caused problems
when using lvm2app even when no activation happens - there are
situations in which we don't need to use mapper/control, but still
need some of the lvm2app functionality. This is also the case for
lvm2-activation systemd generator which just needs to look at the
lvm2 configuration, but it shouldn't touch mapper/control.
Add new lvs segment field 'Monitor' showing 3 states:
"monitored" - LV is monitored by dmeventd.
"not monitored" - LV is currently not being monitored by dmeventd
"" (empty) - LV does not support monitoring, or dmeventd support
is not compiled in.
Support for exclusive activation of snapshots revealed some problems.
When snapshot is created, COW LV is activated first (for clearing) and
then it's transformed into snapshot's COW LV, but it has left the lock
for such LV active in cluster and this lock could not have been removed
from dlm, unless snapshot has been removed within same dlm session.
If the user tried to remove snapshot after rebooting node, the lock was
missing, and COW LV could not have been detached.
Patch modifes the approach in this way:
Always deactivate COW LV for clustered vg after clearing (so it's
activated again via imlicit snapshot activation rule when snapshot is activated).
When snapshot is removed, activate COW LV as independend LV, so the lock
will exist for such LV, but only when the snapshot is active.
Also add test case for testing snapshot removal after cluster reboot.
Patch fixes hidden problem with lvm metadata caching.
When the pretest was made, only the commited data have been cached back
since the call lv_info_by_lvid() triggers mda read operation.
However call of lv_suspend_if_active() also reads precommited metadata.
The problem is visible in this sequence of calls:
vg_write(), suspend_lv(), vg_commit(), resume_lv()
which may end with leaving outdated mda in lvm cache, since vg_write()
drops cached metadata and vg_commit() only transforms precommited
to commited metadata, but in the case of pretesting we have
no precommited mda available so the cache will continue to use
old metadata. This happens, when suspend LV is inactive.
Merging multiple config files together needs to know newest (highest)
timestamp of merged files. Persistent cache file is being used
only in case, the config file is older then .cache file.
Assign fid as the last step before returning VG.
Make the format reader for 'lvm1' and 'pool' equal to 'lvm2' format reader.
It has caused memory corruption to lvmetad as it later calls
destroy_instance() to allocated fid. This patch should fix problems
with crashing test lvmetad-lvm1.sh.
'lvchange' is used to alter a RAID 1 logical volume's write-mostly and
write-behind characteristics. The '--writemostly' parameter takes a
PV as an argument with an optional trailing character to specify whether
to set ('y'), unset ('n'), or toggle ('t') the value. If no trailing
character is given, it will set the flag.
Synopsis:
lvchange [--writemostly <PV>:{t|y|n}] [--writebehind <count>] vg/lv
Example:
lvchange --writemostly /dev/sdb1:y --writebehind 512 vg/raid1_lv
The last character in the 'lv_attr' field is used to show whether a device
has the WriteMostly flag set. It is signified with a 'w'. If the device
has failed, the 'p'artial flag has priority.
Example ("nosync" raid1 with mismatch_cnt and writemostly):
[~]# lvs -a --segment vg
LV VG Attr #Str Type SSize
raid1 vg Rwi---r-m 2 raid1 500.00m
[raid1_rimage_0] vg Iwi---r-- 1 linear 500.00m
[raid1_rimage_1] vg Iwi---r-w 1 linear 500.00m
[raid1_rmeta_0] vg ewi---r-- 1 linear 4.00m
[raid1_rmeta_1] vg ewi---r-- 1 linear 4.00m
Example (raid1 with mismatch_cnt, writemostly - but failed drive):
[~]# lvs -a --segment vg
LV VG Attr #Str Type SSize
raid1 vg rwi---r-p 2 raid1 500.00m
[raid1_rimage_0] vg Iwi---r-- 1 linear 500.00m
[raid1_rimage_1] vg Iwi---r-p 1 linear 500.00m
[raid1_rmeta_0] vg ewi---r-- 1 linear 4.00m
[raid1_rmeta_1] vg ewi---r-p 1 linear 4.00m
A new reportable field has been added for writebehind as well. If
write-behind has not been set or the LV is not RAID1, the field will
be blank.
Example (writebehind is set):
[~]# lvs -a -o name,attr,writebehind vg
LV Attr WBehind
lv rwi-a-r-- 512
[lv_rimage_0] iwi-aor-w
[lv_rimage_1] iwi-aor--
[lv_rmeta_0] ewi-aor--
[lv_rmeta_1] ewi-aor--
Example (writebehind is not set):
[~]# lvs -a -o name,attr,writebehind vg
LV Attr WBehind
lv rwi-a-r--
[lv_rimage_0] iwi-aor-w
[lv_rimage_1] iwi-aor--
[lv_rmeta_0] ewi-aor--
[lv_rmeta_1] ewi-aor--
Move common code for changing activation state from
vgchange and lvchange to one function.
Fix the order of checks - so we always implicitelly
activate snapshots and thin volumes in exclusive mode,
and we do not allow local deactivation for them.
New options to 'lvchange' allow users to scrub their RAID LVs.
Synopsis:
lvchange --syncaction {check|repair} vg/raid_lv
RAID scrubbing is the process of reading all the data and parity blocks in
an array and checking to see whether they are coherent. 'lvchange' can
now initaite the two scrubbing operations: "check" and "repair". "check"
will go over the array and recored the number of discrepancies but not
repair them. "repair" will correct the discrepancies as it finds them.
'lvchange --syncaction repair vg/raid_lv' is not to be confused with
'lvconvert --repair vg/raid_lv'. The former initiates a background
synchronization operation on the array, while the latter is designed to
repair/replace failed devices in a mirror or RAID logical volume.
Additional reporting has been added for 'lvs' to support the new
operations. Two new printable fields (which are not printed by
default) have been added: "syncaction" and "mismatches". These
can be accessed using the '-o' option to 'lvs', like:
lvs -o +syncaction,mismatches vg/lv
"syncaction" will print the current synchronization operation that the
RAID volume is performing. It can be one of the following:
- idle: All sync operations complete (doing nothing)
- resync: Initializing an array or recovering after a machine failure
- recover: Replacing a device in the array
- check: Looking for array inconsistencies
- repair: Looking for and repairing inconsistencies
The "mismatches" field with print the number of descrepancies found during
a check or repair operation.
The 'Cpy%Sync' field already available to 'lvs' will print the progress
of any of the above syncactions, including check and repair.
Finally, the lv_attr field has changed to accomadate the scrubbing operations
as well. The role of the 'p'artial character in the lv_attr report field
as expanded. "Partial" is really an indicator for the health of a
logical volume and it makes sense to extend this include other health
indicators as well, specifically:
'm'ismatches: Indicates that there are discrepancies in a RAID
LV. This character is shown after a scrubbing
operation has detected that portions of the RAID
are not coherent.
'r'efresh : Indicates that a device in a RAID array has suffered
a failure and the kernel regards it as failed -
even though LVM can read the device label and
considers the device to be ok. The LV should be
'r'efreshed to notify the kernel that the device is
now available, or the device should be 'r'eplaced
if it is suspected of failing.
...to not pollute the common and format-independent code in the
abstraction layer above.
The format1 pv_write has common code for writing metadata and
PV header by calling the "write_disks" fn and when rewriting
the header itself only (e.g. just for the purpose of changing
the PV UUID) during the pvchange operation, we had to tweak
this functionality for the format1 case and we had to assign
the PV the orphan state temporarily.
This patch removes the need for this format1 tweak and it calls
the write_disks with appropriate flag indicating whether this is
a PV write call or a VG write call, allowing for metatada update
for the latter one.
Also, a side effect of the former tweak was that it effectively
invalidated the cache (even for the non-format1 PVs) as we
assigned it the orphan state temporarily just for the format1
PV write to pass.
Also, that tweak made it difficult to directly detect whether
a PV was part of a VG or not because the state was incorrect.
Also, it's not necessary to backup and restore some PV fields
when doing a PV write:
orig_pe_size = pv_pe_size(pv);
orig_pe_start = pv_pe_start(pv);
orig_pe_count = pv_pe_count(pv);
...
pv_write(pv)
...
pv->pe_size = orig_pe_size;
pv->pe_start = orig_pe_start;
pv->pe_count = orig_pe_count;
...this is already done by the layer below itself (the _format1_pv_write fn).
So let's have this cleaned up so we don't need to be bothered
about any 'format1 special case for pv_write' anymore.
Add basic support for converting LV into an external origin volume.
Syntax:
lvconvert --thinpool vg/pool --originname renamed_origin -T origin
It will convert volume 'origin' into a thin volume, which will
use 'renamed_origin' as an external read-only origin.
All read/write into origin will go via 'pool'.
renamed_origin volume is read-only volume, that could be activated
only in read-only mode, and cannot be modified.
Reorder activation code to look similar for preload tree and
activation tree.
Its also give much better suppport for device stacking,
since now we also support activation of snapshot which might
be then used for other devices.
A new function (dm_tree_node_force_identical_table_reload) was added to
avoid the suppression of identical table reloads. This allows RAID LVs
to reload the on-disk superblock information that contains which devices
have failed and the bitmaps. If the failed device has returned, this has
the effect of restoring the device and initiating recovery. Without this
patch, the user had to completely deactivate their RAID LV and re-activate
it in order to restore the failed device. Now they simply need to
suspend and resume (which is done by 'lvchange --refresh').
The identical table suppression is only avoided if the LV is not PARTAIL
(i.e. all of it's devices can be seen and read by LVM) and the kernel
status of the array contains failed devices. In other words, the function
will only be called in the case where we may have success in restoring
a failed device in the array.
When there are missing PVs in a volume group, most operations that alter
the LVM metadata are disallowed. It turns out that 'vgimport' is one of
those disallowed operations. This is bad because it creates a circular
dependency. 'vgimport' will complain that the VG is inconsistent and that
'vgreduce --removemissing' must be run. However, 'vgreduce' cannot be run
because it has not been imported. Therefore, 'vgimport' must be one of
the operations allowed to change the metadata when PVs are missing. The
'--force' option is the way to make 'vgimport' happen in spite of the
missing PVs.
If zero metadata copies are used, there's no further recalculation of
PV alignment that happens when adding metadata areas to the PV and
which actually calculates the alignment correctly as a matter of fact.
So fix this for "PV without MDA" case as well.
Before this patch:
[1] raw/~ # pvcreate --dataalignment 8m --dataalignmentoffset 4m
--metadatacopies 1 /dev/sda
Physical volume "/dev/sda" successfully created
[1] raw/~ # pvs -o pv_name,pe_start
PV 1st PE
/dev/sda 12.00m
[1] raw/~ # pvcreate --dataalignment 8m --dataalignmentoffset 4m
--metadatacopies 0 /dev/sda
Physical volume "/dev/sda" successfully created
[1] raw/~ # pvs -o pv_name,pe_start
PV 1st PE
/dev/sda 8.00m
After this patch:
[1] raw/~ # pvcreate --dataalignment 8m --dataalignmentoffset 4m
--metadatacopies 1 /dev/sda
Physical volume "/dev/sda" successfully created
[1] raw/~ # pvs -o pv_name,pe_start
PV 1st PE
/dev/sda 12.00m
[1] raw/~ # pvcreate --dataalignment 8m --dataalignmentoffset 4m
--metadatacopies 0 /dev/sda
Physical volume "/dev/sda" successfully created
[1] raw/~ # pvs -o pv_name,pe_start
PV 1st PE
/dev/sda 12.00m
Also, remove a superfluous condition "pv->pe_start < pv->pe_align" in:
if (pe_start == PV_PE_START_CALC && pv->pe_start < pv->pe_align)
pv->pe_start = pv->pe_align ...
This part of the condition is not reachable as with the PV_PE_START_CALC,
we always have pv->pe_start set to 0 from the PV struct initialisation
(...the pv->pe_start value is just being calculated).
When a device fails, we may wish to replace those segments with an
error segment. (Like when a 'vgreduce --removemissing' removes a
failed device that happens to be a RAID image/meta.) We are then left
with images that we will eventually want to remove or replace.
This patch allows us to pull out these virtual "error" sub-LVs. This
allows a user to 'lvconvert -m -1 vg/lv' to extract the bad sub-LVs.
Sub-LVs with error segments are considered for extraction before other
possible devices so that good devices are not accidentally removed.
This patch also adds the ability to replace RAID images that contain error
segments. The user will still be unable to run 'lvconvert --replace'
because there is no way to address the 'error' segment (i.e. no PV
that it is associated with). However, 'lvconvert --repair' can be
used to replace the image's error segment with a new PV. This is also
the most appropriate way to do it, since the LV will continue to be
reported as 'partial'.
Currently it is impossible to remove a failed PV which has a RAID LV
on it. This patch fixes the issue by replacing the failed PV with an
'error' segment within the affected sub-LVs. Once there is no longer
a RAID LV using the PV, it can be removed.
Most often, it is better to replace a failed RAID device with a spare.
(You can use 'lvconvert --repair <vg>/<LV>' to accomplish that.)
However, if there are no spares in the volume group and none will be
added, it is useful to be able to removed the failed device.
Following patches address the ability to perform 'lvconvert' operations
on RAID LVs that contain sub-LVs composed of 'error' segments.
We have been using 'mirror_region_size' in lvm.conf as the default region
size for RAID logical volumes as well as mirror logical volumes. Since,
"raid" is more inclusive and representative than "mirror", I have changed
the name of this setting. We must still check for the old setting and warn
the user if we are overriding it with the new setting if both happen to be
present.
Instead of check for lv_is_active() for thin pool LV,
query the whole pool via new pool_is_active().
Fixes a problem when we cannot change discards settings
for active pool device where the actual layer for pool
device was inactive, but thin volumes using thin pool
have been active.
This internal function check for active pool device.
For cluster it checks every thin volume,
On the non-clustered VG we need to check just
for presence of -tpool device.
There are currently a few issues with the reporting done on RAID LVs and
sub-LVs. The most concerning is that 'lvs' does not always report the
correct failure status of individual RAID sub-LVs (devices). This can
occur when a device fails and is restored after the failure has been
detected by the kernel. In this case, 'lvs' would report all devices are
fine because it can read the labels on each device just fine.
Example:
[root@bp-01 lvm2]# lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-- 100.00 lv_rimage_0(0),lv_rimage_1(0)
[lv_rimage_0] vg iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg iwi-aor-- /dev/sdb1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
However, 'dmsetup status' on the device tells us a different story:
[root@bp-01 lvm2]# dmsetup status vg-lv
0 1024000 raid raid1 2 DA 1024000/1024000
In this case, we must also be sure to check the RAID LVs kernel status
in order to get the proper information. Here is an example of the correct
output that is displayed after this patch is applied:
[root@bp-01 lvm2]# lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-p 100.00 lv_rimage_0(0),lv_rimage_1(0)
[lv_rimage_0] vg iwi-aor-p /dev/sda1(1)
[lv_rimage_1] vg iwi-aor-- /dev/sdb1(1)
[lv_rmeta_0] vg ewi-aor-p /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
The other case where 'lvs' gives incomplete or improper output is when a
device is replaced or added to a RAID LV. It should display that the RAID
LV is in the process of sync'ing and that the new device is the only one
that is not-in-sync - as indicated by a leading 'I' in the Attr column.
(Remember that 'i' indicates an (i)mage that is in-sync and 'I' indicates
an (I)mage that is not in sync.) Here's an example of the old incorrect
behaviour:
[root@bp-01 lvm2]# lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-- 100.00 lv_rimage_0(0),lv_rimage_1(0)
[lv_rimage_0] vg iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg iwi-aor-- /dev/sdb1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
[root@bp-01 lvm2]# lvconvert -m +1 vg/lv; lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-- 0.00 lv_rimage_0(0),lv_rimage_1(0),lv_rimage_2(0)
[lv_rimage_0] vg Iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg Iwi-aor-- /dev/sdb1(1)
[lv_rimage_2] vg Iwi-aor-- /dev/sdc1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
[lv_rmeta_2] vg ewi-aor-- /dev/sdc1(0) ** Note that all the images currently are marked as 'I' even though it is
only the last device that has been added that should be marked.
Here is an example of the correct output after this patch is applied:
[root@bp-01 lvm2]# lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-- 100.00 lv_rimage_0(0),lv_rimage_1(0)
[lv_rimage_0] vg iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg iwi-aor-- /dev/sdb1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
[root@bp-01 lvm2]# lvconvert -m +1 vg/lv; lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg rwi-a-r-- 0.00 lv_rimage_0(0),lv_rimage_1(0),lv_rimage_2(0)
[lv_rimage_0] vg iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg iwi-aor-- /dev/sdb1(1)
[lv_rimage_2] vg Iwi-aor-- /dev/sdc1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-- /dev/sdb1(0)
[lv_rmeta_2] vg ewi-aor-- /dev/sdc1(0)
** Note only the last image is marked with an 'I'. This is correct and we can
tell that it isn't the whole array that is sync'ing, but just the new
device.
It also works under snapshots...
[root@bp-01 lvm2]# lvs -a -o name,vg_name,attr,copy_percent,devices vg
LV VG Attr Cpy%Sync Devices
lv vg owi-a-r-p 33.47 lv_rimage_0(0),lv_rimage_1(0),lv_rimage_2(0)
[lv_rimage_0] vg iwi-aor-- /dev/sda1(1)
[lv_rimage_1] vg Iwi-aor-p /dev/sdb1(1)
[lv_rimage_2] vg Iwi-aor-- /dev/sdc1(1)
[lv_rmeta_0] vg ewi-aor-- /dev/sda1(0)
[lv_rmeta_1] vg ewi-aor-p /dev/sdb1(0)
[lv_rmeta_2] vg ewi-aor-- /dev/sdc1(0)
snap vg swi-a-s-- /dev/sda1(51201)
If there was a nested mountpoint inside an existing mount path,
blkdeactivate could fail to unmount such a mountpoint as it
needs to deactivate the deepest path first and continue upwards.
For example the simplest reproducer:
[root@rhel6-a ~]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 4G 0 disk
|-vg-lvol0 (dm-2) 253:2 0 32M 0 lvm /mnt/a
`-vg-lvol1 (dm-3) 253:3 0 32M 0 lvm /mnt/a/b
Before this patch:
[root@rhel6-a ~]# blkdeactivate -u
Deactivating block devices:
UMOUNT: unmounting vg-lvol0 (dm-2) mounted on /mnt/a
umount: /mnt/a: device is busy.
(In some cases useful info about processes that use
the device is found by lsof(8) or fuser(1))
UMOUNT: unmounting vg-lvol1 (dm-3) mounted on /mnt/a/b
LVM: deactivating Logical Volume vg/lvol1
(deactivation of vg/lvol0 is skipped as /mnt/a that is on lvol0
can't be unmounted - it still has /mnt/a/b as nested mountpoint!)
With this patch applied:
[root@rhel6-a ~]# blkdeactivate -u
Deactivating block devices:
UMOUNT: unmounting vg-lvol1 (dm-3) mounted on /mnt/a/b
UMOUNT: unmounting vg-lvol0 (dm-2) mounted on /mnt/a
LVM: deactivating Logical Volume vg/lvol0
LVM: deactivating Logical Volume vg/lvol1
===
Also, this patch contains a fix for processing mangled mount paths:
[root@rhel6-a ~]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 4G 0 disk
`-vg-lvol0 (dm-2) 253:2 0 32M 0 lvm /mnt/x y z
[root@rhel6-a ~]# lsblk -r
vg-lvol0 253:2 0 32M 0 lvm /mnt/x\x20y\x20z
(the mount path is mangled with \xNN that is visible in raw
lsblk output only and which is used in blkdeactive as well)
Before this patch:
[root@rhel6-a ~]# blkdeactivate -u
Deactivating block devices:
umount: /mnt/x\x20y\x20z: not found
After this patch applied:
[root@rhel6-a ~]# blkdeactivate -u
Deactivating block devices:
UMOUNT: unmounting vg-lvol0 (dm-2) mounted on /mnt/x\x20y\x20z
LVM: deactivating Logical Volume vg/lvol0
For reseting locale environment into significantly less memory
consuming version 'C' - use LC_ALL instead of LANG since it has
higher priority in locale settings.
Otherwise we may observe whole locale-archive which might be
over 100MB on i.e. Fedora systems locked in memory with
some daemons.
Add log/debug_classes to lvm.conf to allow debug messages to be
classified and filtered at runtime.
The dm_errno field is only used by log_error(), so I've redefined it
for log_debug() messages to hold the message class.
By default, all existing messages appear, but we can add categories that
generate high volumes of data, such as logging all traffic to/from
lvmetad.
We need to call sync_local_dev_names directly as pvscan uses
VG_GLOBAL lock and this one *does not* cause the synchronization
(sync_dev_names) to be called on unlock (VG_GLOBAL is not a real VG):
define unlock_vg(cmd, vol)
do { \
if (is_real_vg(vol)) \
sync_dev_names(cmd); \
(void) lock_vol(cmd, vol, LCK_VG_UNLOCK); \
} while (0)
Without this fix, we end up without udev synchronization for the
pvscan --cache (mainly for -aay that causes the VGs/LVs to be
autoactivated) and also udev synchronization cookies are then left
in the system since they're not managed properly (code before sets
up udev sync cookies, but we have to call dm_udev_wait at least once
after that to do the wait and cleanup).
Before, the pvscan --cache -aay was called on each ADD and CHANGE
uevent (for a device that is not a device-mapper device) and each CHANGE
event (for a PV that is a device-mapper device).
This causes troubles with autoactivation in some cases as CHANGE event
may originate from using the OPTION+="watch" udev rule that is defined
in 60-persistent-storage.rules (part of the rules provided by udev
directly) and it's used for all block devices
(except fd*|mtd*|nbd*|gnbd*|btibm*|dm-*|md* devices). For example, the
following sequence incorrectly activates the rest of LVs in a VG if one
of the LVs in the VG is being removed:
[root@rhel6-a ~]# pvcreate /dev/sda
Physical volume "/dev/sda" successfully created
[root@rhel6-a ~]# vgcreate vg /dev/sda
Volume group "vg" successfully created
[root@rhel6-a ~]# lvcreate -l1 vg
Logical volume "lvol0" created
[root@rhel6-a ~]# lvcreate -l1 vg
Logical volume "lvol1" created
[root@rhel6-a ~]# vgchange -an vg
0 logical volume(s) in volume group "vg" now active
[root@rhel6-a ~]# lvs
LV VG Attr LSize Pool Origin Data% Move Log
Cpy%Sync Convert
lvol0 vg -wi------ 4.00m
lvol1 vg -wi------ 4.00m
[root@rhel6-a ~]# lvremove -ff vg/lvol1
Logical volume "lvol1" successfully removed
[root@rhel6-a ~]# lvs
LV VG Attr LSize Pool Origin Data% Move Log
Cpy%Sync Convert
lvol0 vg -wi-a---- 4.00m
...so the vg was deactivated, then lvol1 removed, and we end up with
lvol1 removed (which is ok) BUT with lvol0 activated (which is wrong)!!!
This is because after lvol1 removal, we need to write metadata to the
underlying device /dev/sda and that causes the CHANGE event to be
generated (because of the WATCH udev rule set on this device) and this
causes the pvscan --cache -aay to be reevaluated.
We have to limit this and call pvscan --cache -aay to autoactivate
VGs/LVs only in these cases:
--> if the *PV is not a dm device*, scan only after proper device
addition (ADD event) and not with any other changes (CHANGE event)
--> if the *PV is a dm device*, scan only after proper mapping
activation (CHANGE event + the underlying PV in a state "just
activated")
If a RAID array is not in-sync, replacing devices should not be allowed
as a general rule. This is because the contents used to populate the
incoming device may be undefined because the devices being read where
not in-sync. The kernel enforces this rule unless overridden by not
allowing the creation of an array that is not in-sync and includes a
devices that needs to be rebuilt.
Since we cannot know the sync state of an LV if it is inactive, we must
also enforce the rule that an array must be active to replace devices.
That leaves us with the following conditions:
1) never allow replacement or repair of devices if the LV is in-active
2) never allow replacement if the LV is not in-sync
3) allow repair if the LV is not in-sync, but warn that contents may
not be recoverable.
In the case where a user is performing the repair on the command line via
'lvconvert --repair', the warning is printed before the user is prompted
if they would like to replace the device(s). If the repair is automated
(i.e. via dmeventd and policy is "allocate"), then the device is replaced
if possible and the warning is printed.
If the lvmcache_info_from_pvid() fails to find valid
info, invoke the lookup by dev, and only in this case
call lvmcache_info_from_pvid() again.
Also check for the result of info and return
error directly, so the NULL is not passed
to lvmcache_get_label().
If we fail to get memory for mutex, hash the mutex
or fail somewhere along pthread function calls
return allocated resources back and unlock vg_lock_map mutex.
Detect failure of dm_pool_strdup() and print error in fail path.
Save one extra strchr call - since we already know the distance
for the '=' character.
Drop stack trace from return after log_error().
When the abort_on_internal_errors is enabled, we aborted prior
the syslog logging output.
Since such fatal error gets level _LOG_FATAL it should
not be blocked by debug_level() check so lets move it further,
to get abort error logged also via syslog.
Tabify
Remove use of asize, unneeded.
Don't initialize lvobj->parent_vgobj to NULL, the object ctor already
zeroed everything on alloc.
Redo call to lvm_lv_snapshot to use the liblvm snapshot implementation
we went with.
Add {}s to silence warning in lv_dealloc.
Rename snapshot function for consistency.
Update WHATS_NEW.
Signed-off-by: Andy Grover <agrover@redhat.com>
Function _ignore_blocked_mirror_devices was not release
allocated strings images_health and log_health.
In error paths it was also not releasing dm_task structure.
Swaped return code of _ignore_blocked_mirror_devices and
use 1 as success.
In _parse_mirror_status use log_error if memory allocation
fails and few more errors so they are no going unnoticed
as debug messages.
On error path always clear return values and free strings.
For dev_create_file use cache mem pool to avoid memleak.
Attempting pvmove on RAID LVs replaces the kernel RAID target with
a temporary pvmove target, ultimately destroying the RAID LV. pvmove
must be prevented on RAID LVs for now.
Use 'lvconvert --replace old_pv vg/lv new_pv' if you want to move
an image of the RAID LV.
In case we don't want to activate, autoactivate or have the
VG/LV read-only. Primarily targeted for the auto_activation_volume_list,
but it makes no harm for other settings (the part of the code
that reads these three settings is shared, but there's no
reason to separate it only for this change).
Rework thin feature detection to support runtime
section to allow to disable them selectively.
New lvm.conf option is born: global/thin_disabled_features
Support swapping of metadata device if the thin pool already
exists. This way it's easy to i.e. resize metadata or their
repair operation.
User may create some empty LV, replace existing metadata
or dump and restore them into bigger LV.
Setting this environment variable will cause a full fallback
to old direct node and symlink management in libdevmapper and lvm2.
It means:
- disabling udev synchronization
(--noudevsync in dmsetup and --noudevsync + activation/udev_sync=0
lvm2 config)
- disabling dm and any subsystem related udev rules
(--noudevrules in dmsetup and activation/udev_rules=0 lvm2 config)
- management of nodes/symlinks under /dev directly by libdevmapper/lvm2
(--verifyudev in dmsetup and activation/verify_udev_operations=1
lvm2 config)
- not obtaining any device list from udev database
(devices/obtain_device_list_from_udev=0 lvm2 config)
Note: we could set all of these before - there's no functional change!
However the DM_DISABLE_UDEV environment variable is a nice shortcut
to make it easier for libdevmapper users so that one can switch off all
of the udev management off at one go directly on the command line,
without a need to modify any source or add any extra switches.
If udev synchronization is disabled by means of --noudevsync
option, we should disable just the synchronization and nothing else.
The udev fallback (verifying udev operations and fixing the
nodes/symlinks if found incorrect) is orthogonal and controlled
by a separate activation/verify_udev_operations configuration option.
Allow restoring metadata with thin pool volumes.
No validation is done for this case within vgcfgrestore tool -
thus incorrect metadata may lead to destruction of pool content.
Configurable settings for thin pool create
if they are not specified on command line.
New supported lvm.conf options are:
allocation/thin_pool_chunk_size
allocation/thin_pool_discards
allocation/thin_pool_zero
Check if target supports discards for chunk sizes,
that are not power of 2 (just multiple of 64K),
and enable it in case it's supported by thin kernel target.
Similar to the way the 'mirror', 'raid1' and 'raid10' segment types set
the number of mirrors to 2 ('-m 1') if the argument is not specified,
here we set the number of stripes to 2 if not given on the command line
when creating a RAID10 LV.
Move common functions for lvcreate and lvconvert.
get_pool_params() - read thin pool args.
update_pool_params() - updates/validates some thin args.
It is getting complicated and even few more things will be
implemented, so to avoid reimplementing things differently
in lvcreate and lvconvert code has been splitted
into 2 common functions that allow some future extension.
This patch is intended to fix bug 825323 - FS turns read-only during a double
fault of a mirror leg and mirrored log's leg at the same time. It only
affects a 2-way mirror with a mirrored log. 3+-way mirrors and mirrors
without a mirrored log are not affected.
The problem resulted from the fact that the top level mirror was not
using 'noflush' when suspending before its "down-convert". When a
mirror image fails, the bios are queue until a suspend is recieved. If
it is a 'noflush' suspend, the bios can be safely requeued in the DM
core. If 'noflush' is not used, the bios must be pushed through the
target and if a device is failed for a mirror, that means issuing an
error. When an error is received by a file system, it results in it
turning read-only (depending on the FS).
Part of the problem was is due to the nature of the stacking involved in
using a mirror as a mirror's log. When an image in each fail, the top
level mirror stalls because it is waiting for a log flush. The other
stalls waiting for corrective action. When the repair command is issued,
the entire stacked arrangement is collapsed to a linear LV. The log
flush then fails (somewhat uncleanly) and the top-level mirror is suspended
without 'noflush' because it is a linear device.
This patch allows the log to be repaired first, which in turn allows the
top-level mirror's log flush to complete cleanly. The top-level mirror
is then secondarily reduced to a linear device - at which time this mirror
is suspended properly with 'noflush'.
Don't use lvmetad in lvm2-monitor.service ExecStop to avoid a systemd issue.
- a systemd design issue while processing dependencies
with socket-based activation that ends up with a hang
- https://bugzilla.redhat.com/show_bug.cgi?id=843587
(also tracker bug https://bugzilla.redhat.com/show_bug.cgi?id=871527)
- not using lvmetad in this case is just a workaround, once the bug
above is resolved, we should enable the lvmetad in that specific case
Remove dependency on fedora-storage-init.service in lvm2 systemd units.
- fedora-storage-init.service and fedora-storage-init-late.service is
going to be separated into respective units that belong to each block
device subsystem:
- mpath + mdraid activated via udev solely
- dmraid with its own dmraid-activation.service unit
- lvm2 with the lvm2-activation-generator to generate the
activation units runtime if lvmetad disabled
(global/use_lvmetad=0 set in lvm.conf) and activation done
via udev+lvmetad if lvmetad enabled (global/use_lvmetad=1 set
in lvm.conf)
Depend on lvm2-lvmetad.socket in lvm2-monitor.service systemd unit.
- as lvm2-monitor uses lvmetad if lvmetad is enabled
Commit 9fd7ac7d03 did not handle mirrors
that contained mirrored logs. This is because the status line of the
mirror does not give an indication of the health of the mirrored log,
as you can see here:
[root@bp-01 lvm2]# dmsetup status vg-lv vg-lv_mlog
vg-lv: 0 409600 mirror 2 253:6 253:7 400/400 1 AA 3 disk 253:5 A
vg-lv_mlog: 0 8192 mirror 2 253:3 253:4 7/8 1 AD 1 core
Thus, the possibility for LVM commands to hang still persists when mirror
have mirrored logs. I discovered this while performing some testing that
does polling with 'pvs' while doing I/O and killing devices. The 'pvs'
managed to get between the mirrored log device failure and the attempt
by dmeventd to repair it. The result was a very nasty block in LVM
commands that is very difficult to remove - even for someone who knows
what is going on. Thus, it is absolutely essential that the log of a
mirror be recursively checked for mirror devices which may be failed
as well.
Despite what the code comment says in the aforementioned commit...
+ * _mirrored_transient_status(). FIXME: It is unable to handle mirrors
+ * with mirrored logs because it does not have a way to get the status of
+ * the mirror that forms the log, which could be blocked.
... it is possible to get the status of the log because the log device
major/minor is given to us by the status output of the top-level mirror.
We can use that to query the log device for any DM status and see if it
is a mirror that needs to be bypassed. This patch does just that and is
now able to avoid reading from mirrors that have failed devices in a
mirrored log.
Addresses: rhbz855398 (Allow VGs to be built on cluster mirrors),
and other issues.
The LVM code attempts to avoid reading labels from devices that are
suspended to try to avoid situations that may cause the commands to
block indefinitely. When scanning devices, 'ignore_suspended_devices'
can be set so the code (lib/activate/dev_manager.c:device_is_usable())
checks any DM devices it finds and avoids them if they are suspended.
The mirror target has an additional mechanism that can cause I/O to
be blocked. If a device in a mirror fails, all I/O will be blocked
by the kernel until a new table (a linear target or a mirror with
replacement devices) is loaded. The mirror indicates that this condition
has happened by marking a 'D' for the faulty device in its status
output. This condition must also be checked by 'device_is_usable()' to
avoid the possibility of blocking LVM commands indefinitely due to an
attempt to read the blocked mirror for labels.
Until now, mirrors were avoided if the 'ignore_suspended_devices'
condition was set. This check seemed to suggest, "if we are concerned
about suspended devices, then let's ignore mirrors altogether just
in case". This is insufficient and doesn't solve any problems. All
devices that are suspended are already avoided if
'ignore_suspended_devices' is set; and if a mirror is blocking because
of an error condition, it will block the LVM command regardless of the
setting of that variable.
Rather than avoiding mirrors whenever 'ignore_suspended_devices' is
set, this patch causes mirrors to be avoided whenever they are blocking
due to an error. (As mentioned above, the case where a DM device is
suspended is already covered.) This solves a number of issues that weren't
handled before. For example, pvcreate (or any command that does a
pv_read or vg_read, which eventually call device_is_usable()) will be
protected from blocked mirrors regardless of how
'ignore_suspended_devices' is set. Additionally, a mirror that is
neither suspended nor blocking is /allowed/ to be read regardless
of how 'ignore_suspended_devices' is set. (The latter point being the
source of the fix for rhbz855398.)
clvmd -d option parsing was not working properly.
clvmd -d 2 (with space) has been ignored because of
'::' used in getopt string, and as failsafe it's been used '1'.
Later this debug_arg has been ignored and debug_opt was used
instead which happend to have value '1'.
Submitted-by: Robert Milasan <rmilasan at suse.com>
Reported-by: Robert Milasan <rmilasan at suse.com>
Use log_warn to print non-fatal warning messages.
Use of log_error would confuse checker for testing
whether proper error has been reported for some real error.
