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.
There were two bugs before when using pvcreate --restorefile together
with data alignment and its offset specified:
- the --dataalignment was always ignored due to missing braces in the
code when validating the divisibility of supplied --dataalignment
argument with pe_start which we're just restoring:
if (pp->rp.pe_start % pp->data_alignment)
log_warn("WARNING: Ignoring data alignment %" PRIu64
" incompatible with --restorefile value (%"
PRIu64").", pp->data_alignment, pp->rp.pe_start);
pp->data_alignment = 0
The pp->data_alignment should be zeroed only if the pe_start is not
divisible with data_alignment.
- the check for compatibility of restored pe_start was incorrect too
since it did not properly count with the dataalignmentoffset that
could be supplied together with dataalignment
The proper formula is:
X * dataalignment + dataalignmentoffset == pe_start
So it should be:
if ((pp->rp.pe_start % pp->data_alignment) != pp->data_alignment_offset) {
...ignore supplied dataalignment and dataalignment offset...
}
This test for LV name restriction check name of device is below 128
chars (which is enforced by dm target).
Thus it should not count with device name.
(Though the test for PATH_MAX size should be probably also added,
but this is runtime test, since theoretically devpath might differ in cluster)
It's unclear why we should prohibit use of -v output.
So reenable (like with other 'display' tools)
But -c -m is really unsupported - return invalid cmd.
Sort order for -C|--columns as with other options,
and use short capital name as the first (as with other options).
Also drop multiple reference for pvs/lvs/vgs, since now
the text for -C is really close to referrence of lvm anyway.
Drop unused passed cmd pointer from function.
TODO:
We have two similar functions (though not identical)
lv_manip.c: for_each_sub_lv()
metadata.c: _lv_each_dependency()
They seem to not always match - we should probably convert
to use only a single function.
Use proper vgmem memory pool for allocation of LV name in the vg
and check if new renamed LV is a valid name.
TODO: validation should really use also VG name, othewise we are not
able to tell "vgname-lvname" will be valid.
Commit 1a832398a7 moved
some code from _pvchange_single() to main pvchange() and
introduced exit code regression as return codes have not
been properly changed, thus pvchange command exited
with '0' exit code, even though it has reported error.
Also there is a missing vg unlock in error path.
Fix it by counting the total number of expected calls before
checking for pvname and also unlock and relase vg when
pv is not found.
For a quick overview of config when debugging and to quickly check
which values are different from defaults and which are not defined
in the config and for which defaults are used.
When lvm2 command works with clvmd and uses locking in wrong way,
it may 'leak' certain file descriptors in opened (incorrect) state.
dev_cache_exit then destroys memory pool of cached devices, while
_open_devices list in dev-io.c was still referencing them if they
were still opened.
Patch properly calls _close() function to 'self-heal' from this
invalid state, but it will report internal error (so execution
with abort_on_internal_error causes immediate death). On the
normal 'execution', error is only reported, but memory state is
corrected, and linked list is not referencing devices from
released mempool.
For crash see: https://bugzilla.redhat.com/show_bug.cgi?id=1073886
Smallest supported size for swap device is 40KB, however current
test skipped devices smaller then 4096 sectors (2MB).
Since page is in bytes, convert it to sectors before comparing
with device size (in sectors).
pvscan --uuid was broken since it was using only 128 char buffers
without checking any write size, so any longer device path leads to
crash.
Also ansure format is properly aligned into columns with this option.
The PV size is displayed in sectors, not kilobytes
for 'pvdisplay -c'
Signed-off-by: Thomas Fehr <fehr@suse.de>
Acked-by: Hannes Reinecke <hare@suse.de>
Instead of sending repeatedly LOCAL_SYNC commands to clvmds
like 'lvs', rememeber the last sent commmand, and if there was no other
clvmd command, drop this redundant SYNC call message.
The problem has started with commit:
56cab8cc03
This introduced correct synchronisation of name, when user requests to know
open_count (needs to wait for udev), however it is also executed for
read-only cases like 'lvs' command.
For now implement very simple solution, which is only monitoring
outgoing clvmd command, and when sequence of LOCAL sync names are
recognized, they are skipped automatically.
TODO:
Future solution might move this variable info 'cmd_context' and
use 'needs_sync' flag also i.e. in file locking code.
When the backup is disabled, avoid testing backup presence.
This only leads to errors being logged in debug trace and the missing
backup can't be fixed, since it's disabled.
Check whether lvm dumpconfig --mergedconfig is used only
with --type current (where we're merging current config and
the config supplied on command line). With other types
the config was merged, but it was thrown away since we're
generating other type of config anyway. This lead to a memleak.
Error out if --mergedconfig is used with anything else than
--type current (or without specifying --type in which case
the --type current is used by default).
Do not allow conversion of too small LV into a COW snapshot device.
Without this patch snapshot target is generating these kernel
messages before creation fails:
attempt to access beyond end of device
dm-9: rw=16, want=8, limit=2
attempt to access beyond end of device
...
device-mapper: table: 253:11: snapshot: Failed to read snapshot metadata
device-mapper: ioctl: error adding target to table
device-mapper: reload ioctl on failed: Input/output error
Usage of origin as a snapshot 'COW' volume is unsupported.
Without this test lvm2 is able to generate this ugly internal error message.
To test this:
lvcreate -L1 -n lv1 vg
lvcreate -L1 -n lv2 -s vg/lv1
lvcreate -L1 -n lv3 vg
lvconvert -s vg/lv3 vg/lv1
Internal error: LVs (5) != visible LVs (1) + snapshots (1) + internal LVs (0) in VG vg
Users can create several profiles for how the tools report
the output very easily and then just use
<lvm reporting command> --profile <report_profile_name>
This prevents numerous VG refreshes on each "pvscan --cache -aay" call
if the VG is found complete. We need to issue the refresh only if the PV:
- is new
- was gone before and now it reappears (device "unplug/plug back" scenario)
- the metadata has changed
Let's do this the other way round - this makes more logic than commit b995f06.
So let's allow empty values for global/thin_disabled_features where
such an empty value now means "none of this features are disabled".
The empty pool is also the pool which has yet queued list of messages
and transaction_id == 1.
Problem is exposed when pool is created inactive.
lvcreate -L10 -T vg/pool -an
lvcreate -V10 -T vg/pool
When pool_has_message() is queried with NULL lv and 0 device_id
it should just return 'true' when there is any message queued.
So it needs to return negative value dm_list_empty().
Since there is no user for this code path in code currently,
this bug has not been triggered.
This patch will releases allocated private resources from
startup. Needs previous dm_zalloc patch to ensure unset
private pointer is NULL.
TODO: check on real cluster.
Based on patch:
https://www.redhat.com/archives/lvm-devel/2014-March/msg00015.html
The CPPFunction typedef (among others) have been deprecated in favour of
specific prototyped typedefs since readline 4.2 (circa 2001).
It's been working since because compatibility typedefs have been in
place until they where removed in the recent readline 6.3 release.
Switch to the new style to avoid build breakage.
But also add full backward compatibility with define.
Signed-off-by: Gustavo Zacarias <gustavo zacarias com ar>
If the PV label is lost (e.g. by doing a dd on the device), call
"systemd-run pvscan --cache <major>:<minor>" in 69-dm-lvm-metad.rules
to inform lvmetad about this state.
The reason for this is that ENV{SYSTEMD_WANTS}="lvm2-pvscan@<major>:<minor>"
logic will not cause the pvscan to be fired in this case since this works
only on proper device addition/removal cycle - the lvm2-pvscan service's
ExecStop is called only on proper REMOVE event - the service is bound to
device existence. Hence we need pvscan call via systemd-run (that
instantiates a quick transient service just to call the command).
See also https://bugzilla.redhat.com/show_bug.cgi?id=1063813.
Code uses target driver version for better estimation of
max size of COW device for snapshot.
The bug can be tested with this script:
VG=vg1
lvremove -f $VG/origin
set -e
lvcreate -L 2143289344b -n origin $VG
lvcreate -n snap -c 8k -L 2304M -s $VG/origin
dd if=/dev/zero of=/dev/$VG/snap bs=1M count=2044 oflag=direct
The bug happens when these two conditions are met
* origin size is divisible by (chunk_size/16) - so that the last
metadata area is filled completely
* the miscalculated snapshot metadata size is divisible by extent size -
so that there is no padding to extent boundary which would otherwise
save us
Signed-off-by:Mikulas Patocka <mpatocka@redhat.com>
While stripe size is twice the physical extent size,
the original code will not reduce stripe size to maximum
(physical extent size).
Signed-off-by: Zhiqing Zhang <zhangzq.fnst@cn.fujitsu.com>
When read-only snapshot was created, tool was skipping header
initialization of cow device. If it happened device has been
already containing header from some previous snapshot, it's
been 'reused' for a newly created snapshot instead of being cleared.
To make "lvm dumpconfig --type default" output to be usable like any
other config, we need to comment out lines that have no default value
defined. Otherwise, we'd have the output with config options
with blank or zero values which is not the same as when the value
is not defined! And such configuration can't be feed into lvm again
without further edits. So let's fix this.
Currently this covers these configuration options exactly:
devices/loopfiles
devices/preferred_names
devices/filter
devices/global_filter
devices/types
allocation/cling_tag_list
global/format_libraries
global/segment_libraries
activation/volume_list
activation/auto_activation_volume_list
activation/read_only_volume_list
activation/mlock_filter
metadata/dirs
metadata/disk_areas
metadata/disk_areas/<disk_area>
metadata/disk_areas/<disk_area>/start_sector
metadata/disk_areas/<disk_area>/size
metadata/disk_areas/<disk_area>/id
tags/<tag>
tags/<tag>/host_list
Reorder detection of cmirrord. Now if cmirrord is not
running, target will not try to load kernel log module,
for communication with cmirrord.
Whole check for attrs now also happens just once.
Test raid10 availability as a target feature (instead of doing
it in all the places where raid10 should be checked).
TODO: activation needs runtime validation - so metadata with raid10
are skipped from activation in user-friendly way in lvm2.
Parsing vg structure during supend/commit/resume may require a lot of
memory - so move this into vg_write.
FIXME: there are now multiple cache layers which our doing some thing
multiple times at different levels. Moreover there is now different
caching path with and without lvmetad - this should be unified
and both path should use same mechanism.
Remove 'skip' argument passed into the function.
We always used '0' - as this is the only supported
option (-K) and there is no complementary option.
Also add some testing for behaviour of skipping.
We already have /dev/disk/by-id/dm-uuid-... (which encompasses the
VG UUID and LV UUID in case of LVs since the mapping's UUID is
VG+LV UUID together) and /dev/disk/by-id/dm-name-... (which encompasses
the VG and LV name in case of LVs).
This patch addds /dev/disk/by-id/lvm-pv-uuid-<PV_UUID> that completes
this scheme and makes navigation a bit easier using PV UUIDs since
one can navigate using PV UUIDs only and there's no need to do extra
PV UUID <--> kernel name matching (the PV UUID is stable across reboots).
This may come in handy in various scripts.
Since we already have the PV UUID stored in udev database (as a result
of blkid call - returned in ID_FS_UUID blkid's variable), this operation
is very cheap indeed, just creating the extra one symlink.
There are typically 2 functions for the more advanced segment types that
deal with parameters in lvcreate.c: _get_*_params() and _check_*_params().
(Not all segment types name their functions according to this scheme.)
The former function is responsible for reading parameters before the VG
has been read. The latter is for sanity checking and possibly setting
parameters after the VG has been read.
This patch adds a _check_raid_parameters() function that will determine
if the user has specified 'stripe' or 'mirror' parameters. If not, the
proper number is computed from the list of PVs the user has supplied or
the number that are available in the VG. Now that _check_raid_parameters()
is available, we move the check for proper number of stripes from
_get_* to _check_*.
This gives the user the ability to create RAID LVs as follows:
# 5-device RAID5, 4-data, 1-parity (i.e. implicit '-i 4')
~> lvcreate --type raid5 -L 100G -n lv vg /dev/sd[abcde]1
# 5-device RAID6, 3-data, 2-parity (i.e. implicit '-i 3')
~> lvcreate --type raid6 -L 100G -n lv vg /dev/sd[abcde]1
# If 5 PVs in VG, 4-data, 1-parity RAID5
~> lvcreate --type raid5 -L 100G -n lv vg
Considerations:
This patch only affects RAID. It might also be useful to apply this to
the 'stripe' segment type. LVM RAID may include RAID0 at some point in
the future and the implicit stripes would apply there. It would be odd
to have RAID0 be able to auto-determine the stripe count while 'stripe'
could not.
The only draw-back of this patch that I can see is that there might be
less error checking. Rather than informing the user that they forgot
to supply an argument (e.g. '-i'), the value would be computed and it
may differ from what the user actually wanted. I don't see this as a
problem, because the user can check the device count after creation
and remove the LV if they have made an error.
When clustered VG is available in the system but we don't have
clustering set up for whatever reason, the lvm2-monitor scripts should
not fail completely just because these clustered VGs are skipped during
vgs/vgchange calls in lvm2-monitor initscript/systemd unit.
When the activation units are generated if use_lvmetad=0 (no
autoactivation), use --ignoreskippedcluster option for vgchange calls
since the cluster with cLVM is set up by separate units.
This avoids a situation in which the generated activation units are
improperly in failed state just because of the vgchange return value
when clustered VGs are encountered while the activation of non-clustered
VGs does proceed normally.
lv_active_change will enforce proper activation.
Modification of activation was wrong and lead to misuse of
autoactivation. Fix allows to use proper local exclusive activation,
while the removed code turned this into just exclusive
activation (losing required local property).
lvm2_cluster_activation_red_hat.service.in -> lvm2_cluster_activation_systemd_red_hat.service.in
lvm2_clvmd_red_hat.service.in -> lvm2_clvmd_red_hat.service.in
Edit lvm2-cluster-activation reference on cmirror - take new
lvm2-cmirrord.service, it was just cmirrord(.service) before
as the old initscript was used in compatibility mode.
Also, use WantedBy=multi-user.target instead of sysinit.target
in lvm2-cluster-activation.service.
The commit splits original clvmd service in two new native services
for systemd enabled systems while original init scripts remain unaltered.
New systemd native services:
1) clvmd daemon itself (lvm2_clvmd_red_hat.service.in)
2) (de)activation of clustered VGs (lvm2_cluster_activation_red_hat.service.in)
There're several reasons to split it. First, there's no support for conditional
stop in systemd and AFAIK they don't plan to support it. In other words:
if the deactivation fails for some reason, systemd doesn't care and will simply
kill all remaining processes in original cgroup (by default). Killing the
remaining procs can be suppressed however it doesn't solve the following problem:
You can't repeat the stop command of a failed service. The repeated stop command
is simply not propagated to the service in a failed state. You would have to start
and then try to stop the service again. Unfortunately, this can't be done while
the daemon is still running (and we need the daemon to stay active until all
clustered VGs are deactivated properly).
In a separated setup we need only to restart the failed activation service and
that's fine.
No need to fork lvmetad when running under systemd.
Also, the "lvmetad -R" support has been removed in lvm2 v2.02.98
so remove the ExecReload line that called it on "systemctl reload".
The libblkid can detect DM_snapshot_cow signature and when creating
new LVs with blkid wiping used (allocation/use_blkid_wiping=1 lvm.conf
setting and --wipe y used at the same time - which it is by default).
Do not issue any prompts about this signature when new LV is created
and just wipe it right away without asking questions. Still keep the
log in verbose mode though.
Avoid use of external origin with size unaligned/incompatible with
thin pool chunk size, since the last chunk is not correctly provisioned
when it is overwritten.
Avoid starting conversion of the LV to the thin pool and thin volume
at the same time. Since this is mostly a user mistake, do not try
to just convert to one of those type, since we cannot assume if the
user wanted LV to become thin volume or thin pool.
Before the fix tool reported pretty strange internal error:
Internal error: Referenced LV lvol1_tdata not listed in VG mvg.
Fixed output:
lvconvert --thinpool lvol0 -T mvg/lvol0
Can't use same LV mvg/lvol0 for thin pool and thin volume.
When lvm2 command forks, it calls reset_locking(),
which as an unwanted side effect unlinked lock file from filesystem.
Patch changes the behavior to just close locked file descriptor
in children - so the lock is being still properly hold in the parent.
Test LVM_LVMETAD_PIDFILE for pid for lvm command.
Fix WHATS_NEW envvar name usage
Fix init order in prepare_lvmetad to respect set vars
and avoid clash with system settings.
Update test to really test the 'is running' message.
Comparing for available feature missed the code path, when
maj is already bigger.
The bug would be only hit in the case, thin pool target would have
increased major version.
When thin volume is using external origin, current thin target
is not able to supply 'extended' size with empty pages.
lvm2 detects version and disables extension of LV past the external
origin size in this case.
Thin LV could be however still reduced and extended freely bellow
this size.
We need both offset and length when trying to wipe detected signatures.
The libblkid can fail so it's good to have an error message issued for
this state instead of being silent (libblkid does not issue any error
messages here). We just issued "stack" here before but that was not
quite useful if some error occurs...
Only flag thin LV for no scanning in udev if this LV is about
to be wiped. This happens only in case the thin LV's pool was not
created with zeroing of the new blocks enabled.
DO NOT USE LVMETAD IF YOU HAVE ANY LVM1-FORMATTED PVS.
You may continue to use it without lvmetad, but do please schedule
an upgrade to the lvm2 format (with 'vgconvert').
Sending the original LVM1 formatted metadata to lvmetad is breaking
assumptions made by the code, so I am marking the format as obsolete for
now and no longer sending it to lvmetad.
This means that if you are using lvmetad, lvm1 volumes will usually
appear invisible - though not always: it depends on exactly what
sequence of commands you run!
The current situation is not satisfactory.
We'll either fix lvmetad and reenable this or we'll fix the code to
issue appropriate warning messages when lvm1 PVs are encountered
to avoid accidents.
(The latest unfixed problem is that lvmetad assumes metadata sequence
numbers exist and always increase - but the lvm1 format does not define
or store any sequence number, confusing both the daemon and client
when default values get passed to-and-fro.)
If a PV in an existing VG becomes orphaned (with 'pvcreate -ff', for
example) the VG struct cached against its vginfo must be invalidated.
This is because the struct device it references no longer contains
the PV label so becomes incorrect.
This triggers the error:
Internal error: PV $dev unexpectedly not in cache.
when the PV from the cached VG metadata is subsequently looked up
in the cache.
Bug introduced in 2.02.87 by commit 7ad0d47c3c
("Cache and share generated VG structs").
Before:
lvm> pvs
PV VG Fmt Attr PSize PFree
/dev/loop3 vg12 lvm2 a-- 28.00m 28.00m
/dev/loop4 vg12 lvm2 a-- 28.00m 28.00m
lvm> pvcreate -ff /dev/loop3
Really INITIALIZE physical volume "/dev/loop3" of volume group "vg12" [y/n]? y
WARNING: Forcing physical volume creation on /dev/loop3 of volume group "vg12"
Physical volume "/dev/loop3" successfully created
lvm> pvs
Internal error: PV /dev/loop3 unexpectedly not in cache.
PV VG Fmt Attr PSize PFree
/dev/loop3 vg12 lvm2 a-- 28.00m 28.00m
/dev/loop3 lvm2 a-- 32.00m 32.00m
/dev/loop4 vg12 lvm2 a-- 28.00m 28.00m
After:
lvm> pvs
PV VG Fmt Attr PSize PFree
/dev/loop3 vg12 lvm2 a-- 28.00m 28.00m
/dev/loop4 vg12 lvm2 a-- 28.00m 28.00m
lvm> pvcreate -ff /dev/loop3
Really INITIALIZE physical volume "/dev/loop3" of volume group "vg12" [y/n]? y
WARNING: Forcing physical volume creation on /dev/loop3 of volume group "vg12"
Physical volume "/dev/loop3" successfully created
lvm> pvs
PV VG Fmt Attr PSize PFree
/dev/loop3 lvm2 a-- 32.00m 32.00m
/dev/loop4 vg12 lvm2 a-- 28.00m 28.00m
unknown device vg12 lvm2 a-m 28.00m 28.00m
When using filters for the pvscan --cache (the global_filter),
there's a difference between:
pvscan --cache -aay /dev/block/<major>:<minor>
and
pvscan --cache -aay <major>:<minor> (or --major <major> --minor <minor>)
In the first case, we need to be sure to have an exact matching line
in the filter for the device to be used, no aliases are considered
So for example even if we have accept rule for "/dev/sda" present,
this won't apply for "/dev/block/8:0" even though it's the same device!
This is because we're comparing the path used on command line directly
with the path written in the rule.
For the second one, any alias mentioned in the filter will apply
as we're comparing the major and minor pair, not looking at actual
device names - so any alias mentioned in the rules will suffice for
the filtering rule to apply.
For the global_filter to be properly used, we need to call the
second one in the lvm2-pvscan@.service - nobody is able to tell
what value of major:minor the kernel assignes next time, hence
this bug makes the use of global_filter quite unusable!
This reverts commit 24639be558.
Ok - seems we could be here a bit too active - and we
may remove devices which are unsuable for reasons we are not
aware of - thus taking down whole device could be way to big hammer.
So we still need some solution to recover from failing preload
and activation - but it needs more tunning.
When activation fails - we may leak large tree of partially loaded
devices in the dm table (i.e. failure in snapshot activation)
The best we can do here is try to deactivate whole device and
remove as much inactive table entries as we can.
When LV is scanned for its dependencies - scan also origin's snapshots,
and thin external origins.
So if any PV from snapshot or external origin device is missing - lvm2 will
avoid trying to activate such device.
When the device is inserted in dev_name_confirmed() stat() is
called twice as _insert() has it's own stat() call.
Extend _insert() parameter with struct stat* - which could be used
if it has been just obtained. When NULL is passed code is
doing its own stat() call as before.
Use internal type by default for thin provisioning.
If user is not interested in thin provisiong and doesn't
have thin provisining supporting tools installed,
configure will just print warning at the end of configure
process about limited support.
Boolean algebra changes for process_each_lv_in_vg().
1st.
Drop process_lv variable since it's not needed.
2nd.
process_lv was always initilized to 0 - so the condition was always true.
It the condition (!tags_supplied && !lvargs_supplied) evaluates as "true",
process_all is already set to 1, so skip vg tags evaluation.
3rd.
Move check for matching lv name in the front of lv tags check
since this check can't be skipped for lvargs_matched counter.
If this filter evaluates to true, skip lv tags evaluation.
Thin kernel target 1.9 still does not support online resize of
thin pool metadata properly - so disable it with expectation
for much higher version - and reenable after fixing kernel.
Since activation takes only read-lock, there could be
multiple activation running in parallel.
So instead of checking before taking any real lock,
let the locking resolve the problem and just
detect if the reason for failure has been remote
exlusive activation.
It should be also faster, since each activation does
not need to do explicit lock query.
The PIE and RELRO compiler/linker options can be used to produce a code
some techniques applied that makes the code more immune to some attacks:
- PIE (Position Independent Executable). It can make use of the ASLR
(Address Space Layout Randomization) provided by kernel to avoid
static locations for .text regions of executables (this is the 'pie'
compiler and linker option)
- RELRO (Relocation Read-Only). This prevents overwrite attacks of
the GOT (Global Offset Table) and PLT (Procedure Lookup Table)
used for relocations by making it read-only after all relocations
are resolved (these are the 'relro' and 'now' linker options) -
hence all symbols are resolved at the very start so there's no
need for those tables to be writeable later.
These compiler/linker options are now used by default for daemons
if the compiler/linker supports it.
In the case we have a dir with multiple objects and for
an individual object file we need special define -
allow to define it without adding extra rules.
To ensure dmeventd.o compilation will use EXTRA_FLAGS:
CFLAGS_dmeventd.o += $(EXTRA_FLAGS)
Then it's better to use:
dmeventd.o: CFLAGS += $(EXTRA_FLAGS)
At the end of lvconvert --snapshot with an active origin, the origin
gets reloaded.
Commit 57c0f72b1d ("lvconvert: use
_reload_lv on more places") accidentally replaced this with a snapshot
LV reload (which does nothing because only the origin is active).
Make it easier to run a live lvmetad in debugging mode and
to avoid conflicts if multiple test instances need to be run
alongside a live one.
No longer require -s when -f is used: use built-in default.
Add -p to lvmetad to specify the pid file.
No longer disable pidfile if -f used to run in foreground.
If specified socket file appears to be genuine but stale, remove it
before use.
On error, only remove lvmetad socket file if created by the same
process. (Previous code removes socket even while a running instance
is using it!)
If using lv/vgchange --sysinit -aay and lvmetad is enabled, we'd like to
avoid the direct activation and rely on autoactivation instead so
it fits system initialization scripts.
But if we're calling lv/vgchange --sysinit -aay too early when even
lvmetad service is not started yet, we just need to do the direct
activation instead without printing any error messages (while
trying to connect to lvmetad and not finding its socket).
This patch adds two helper functions - "lvmetad_socket_present" and
"lvmetad_used" which can be used to check for this condition properly
and avoid these lvmetad connections when the socket is not present
(and hence lvmetad is not yet running).
Revert 4777eb6872 which put
target_present check into init_snapshot_merge(). However
this function is also used when parsing metadata. So we would
get this present test performed even when target is not really
needed. So move this target_present test directly into lvconvert.
Fix buggy usage of "" (empty string) as a numerical string
value used for sorting.
On intel 64b platform this was typically resolve
as 0xffffff0000000000 - which is already 'close' to
UINT64_MAX which is used for _minusone64.
On other platforms it might have been giving
different numbers depends on aligment of strings.
Use proper &_minusone64 for sorting value when the reported
value is NUM.
Note: each numerical value needs to be thought about if it needs
default value &_zero64 or &_minusone64 since for cases, were
value of zero is valid, sorting should not be mixing entries
together.
Add wrapper function for dm_report_field_set_value() which returns void
and return 1, so the code could be shorter.
Add wrapper function for percent display _field_set_percent().
There's a tiny race when suspending the device which is part
of the refresh because when suspend ioctl is performed, the
dm kernel driver executes (do_suspend and dm_suspend kernel fn):
step 1: a check whether the dev is already suspended and
if yes it returns success immediately as there's
nothing to do
step 2: it grabs the suspend lock
step 3: another check whether the dev is already suspended
and if found suspended, it exits with -EINVAL now
The race can occur in between step 1 and step 2. To prevent
premature autoactivation failure, we're using a simple retry
logic here before we fail completely. For a complete solution,
we need to fix the locking so there's no possibility for suspend
calls to interleave each other to cause this kind of race.
This is just a workaround. Remove it and replace it with proper
locking once we have that in!
Failures in the temporary mirror used when up-converting cause dmeventd
to issue 'lvconvert --repair' on the sub-LV, <lv_name>_mimagetmp_?. The
'lvconvert' command refuses to deal with this sub-LV outright - it
expects to be given the name of the top-level LV. So, just like we do
with mirrored logs, we strip-off the portion of the name that is not
the top-level LV and issue the command on the top-level LV instead.
This patch fixes mostly cluster behavior but also updates
non-cluster reaction where calls like 'lvchange -aln'
lead to incorrect errors for some segment types.
Fix the implicit activation rules where some segment types could
be activated only in exclusive mode in cluster.
lvm2 command was not preserver 'local' property and incorrectly
converted local activations in to plain exclusive, so the local
activation could have activate volumes exclusively, but remotely.
If the volume_list filters out volume from activation,
it is still success result for this function.
Change the error message back to verbose level.
Detect if the volume is active localy before zeroing,
so we report error a bit later for cases, where volume
could not be activated because it doesn't pass through volume
list (but user still could create volume when he disables
zeroing)
Correct return code of activate_lv_excl().
Function is not supposed to return activation state of
activated volume, but return code of the operation.
Since i.e. when activation filter is allowing to activate
volume on current system, it is still success even though
no volume is activated.
MD can directly create partition devices without a need to run
an extra kpartx or partprobe call. We need to react to this event in
a different way as for bare MD devices - we need to handle the ADD event
for KERNEL=="md[0-9]*p[0-9]*" kernel name and trigger the LVM scanning
to update lvmetad to trigger autoactivation and so on...
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1023250
There is a problem with the way mirrors have been designed to handle
failures that is resulting in stuck LVM processes and hung I/O. When
mirrors encounter a write failure, they block I/O and notify userspace
to reconfigure the mirror to remove failed devices. This process is
open to a couple races:
1) Any LVM process other than the one that is meant to deal with the
mirror failure can attempt to read the mirror, fail, and block other
LVM commands (including the repair command) from proceeding due to
holding a lock on the volume group.
2) If there are multiple mirrors that suffer a failure in the same
volume group, a repair can block while attempting to read the LVM
label from one mirror while trying to repair the other.
Mitigation of these races has been attempted by disallowing label reading
of mirrors that are either suspended or are indicated as blocking by
the kernel. While this has closed the window of opportunity for hitting
the above problems considerably, it hasn't closed it completely. This is
because it is still possible to start an LVM command, read the status of
the mirror as healthy, and then perform the read for the label at the
moment after a the failure is discovered by the kernel.
I can see two solutions to this problem:
1) Allow users to configure whether mirrors can be candidates for LVM
labels (i.e. whether PVs can be created on mirror LVs). If the user
chooses to allow label scanning of mirror LVs, it will be at the expense
of a possible hang in I/O or LVM processes.
2) Instrument a way to allow asynchronous label reading - allowing
blocked label reads to be ignored while continuing to process the LVM
command. This would action would allow LVM commands to continue even
though they would have otherwise blocked trying to read a mirror. They
can then release their lock and allow a repair command to commence. In
the event of #2 above, the repair command already in progress can continue
and repair the failed mirror.
This patch brings solution #1. If solution #2 is developed later on, the
configuration option created in #1 can be negated - allowing mirrors to
be scanned for labels by default once again.
Add LV_TEMPORARY flag for LVs with limited existence during command
execution. Such LVs are temporary in way that they need to be activated,
some action done and then removed immediately. Such LVs are just like
any normal LV - the only difference is that they are removed during
LVM command execution. This is also the case for LVs representing
future pool metadata spare LVs which we need to initialize by using
the usual LV before they are declared as pool metadata spare.
We can optimize some other parts like udev to do a better job if
it knows that the LV is temporary and any processing on it is just
useless.
This flag is orthogonal to LV_NOSCAN flag introduced recently
as LV_NOSCAN flag is primarily used to mark an LV for the scanning
to be avoided before the zeroing of the device happens. The LV_TEMPORARY
flag makes a difference between a full-fledged LV visible in the system
and the LV just used as a temporary overlay for some action that needs to
be done on underlying PVs.
For example: lvcreate --thinpool POOL --zero n -L 1G vg
- first, the usual LV is created to do a clean up for pool metadata
spare. The LV is activated, zeroed, deactivated.
- between "activated" and "zeroed" stage, the LV_NOSCAN flag is used
to avoid any scanning in udev
- betwen "zeroed" and "deactivated" stage, we need to avoid the WATCH
udev rule, but since the LV is just a usual LV, we can't make a
difference. The LV_TEMPORARY internal LV flag helps here. If we
create the LV with this flag, the DM_UDEV_DISABLE_DISK_RULES
and DM_UDEV_DISABLE_OTHER_RULES flag are set (just like as it is
with "invisible" and non-top-level LVs) - udev is directed to
skip WATCH rule use.
- if the LV_TEMPORARY flag was not used, there would normally be
a WATCH event generated once the LV is closed after "zeroed"
stage. This will make problems with immediated deactivation that
follows.
The blkdeactivate script iterates over the list of devices if they're
given as an argument and it tries to umount/deactivate them one by one.
This iteration failed to proceed if any of the umount/deactivation
was unsuccessful - there was a missing "shift" call to move to the
next argument (device) for processing. As a result of this, the same
device was tried again and again, causing an endless loop, never
proceeding to the next device given.
When using ENV{SYSTEMD_WANTS}=lvm2-pvscan@... to instantiate a service
for lvmetad scan when the new PV appears in the system, the service
is started and executed. However, to track device removal, we need
to bind it (the "BindsTo" systemd directive) to a certain .device
systemd unit.
In default systemd setup, the device is tracked by it's name and
sysfs path (there's normally a sysfs path .device systemd unit for
a device and then the device name .device unit as an alias for it).
Neither of these two is useful for lvmetad update as we need to bind
it to device's <major>:<minor> pair.
The /dev/block/<major>:<minor> is the essential symlink under /dev
that exists for each block device (created by default udev rules
provided by udev directly). So let's use this as an alias for
the device's .device unit as well by means of "ENV{SYSTEMD_ALIAS}"
declaration within udev rules which systemd understands (this will
create a new alias "dev-block-<major>:<minor>.device".
Then we can easily bind the "dev-block-<major>:<minor>" device
systemd unit with instantiated lvm2-pvscan@<major>:<minor>.service.
So once the device is removed from the systemd, the
lvm-pvscan@<major>:<minor>.service executes it's ExecStop action
(which in turn notifies lvmetad about the device being gone).
This completes the udev-systemd-lvmetad interaction then.
Before, pvscan recognized either:
pvscan --cache --major <major> --minor <minor>
or
pvscan --cache <DevicePath>
When the device is gone and we need to notify lvmetad about device
removal, only --major/--minor works as we can't translate DevicePath
into major/minor pair anymore. The device does not exist in the system
and we don't keep DevicePath index in lvmetad cache to make the
translation internally into original major/minor pair. It would be
useless to keep this index just for this one exact case.
There's nothing bad about using "--major <major> --minor <minor>",
but it makes our life a bit harder when trying to make an
interconnection with systemd units, mainly with instantiated services
where only one and only one arg can be passed (which is encoded in the
service name).
This patch tries to make this easier by adding support for recognizing
the "<major>:<minor>" as a shortcut for the longer form
"--major <major> --minor <minor>". The rule here is simple: if the argument
starts with "/", it's a DevicePath, otherwise it's a <major>:<minor> pair.
There is no point eating stderr for these commands. In fact the
redirect causes confusion and hurts dubugging.
Also reword an error message if the pvs command fails so as not be
certain that a device is not a PV. Coupled with removing the stderr
redirect this will improve the user experience in the face of errors.
The new lvm2-pvscan@.service is responsible for on-demand execution
of "pvscan --cache --activate ay" which causes lvmetad to be
updated and LVM activation done if the VG is complete.
Also, use udev-systemd mechanism to instantiate the job as the
lvm2-pvscan@$devnode.service on each newly appeared PV in the system.
This prevents the background job to be killed (that would happen
if it was directly forked from udev rule - this behaviour is seen
in recent versions of udev with the help of systemd that can track
detached processes - the detached process would still be in the same
cgroup).
To enable this official udev-systemd protocol for instantiating
background jobs, use new --enable-udev-systemd-background-jobs
configure switch (it's disabled by default). This option is highly
recommended wherever systemd is used!
Prohibit conversion of pool device with active thin volumes.
Properly restore active states only for active thin pool volume.
Use new LV_NOSCAN when converting volume into thin pool's metadata.
This patch reinstates the lv_info call to check for open count of
the LV we're removing/deactivating - this was changed with commit 125712b
some time ago and we relied on the ioctl retry logic deeper in the libdm
while calling the exact 'remove' ioctl.
However, there are still some situations in which it's still required to
check for open count before we do any 'remove' actions - this mainly
applies to LVs which consist of several sub LVs, like it is for
virtual snapshot devices.
The commit 1146691 fixed the issue with ordering of actions during
virtual snapshot removal while the snapshot is still open. But
the check for the open status of the snapshot is still prone to
marking the snapshot as in use with an immediate exit even though
this could be a temporary asynchronous open only, most notably
because of udev and its WATCH udev rule with accompanying scans
for the event which is asynchronous. The situation where this crops
up most often is when we're closing the LV that was open for read-write
and then calling lvremove immediately.
This patch reinstates the original lv_info call for the open status
of the LV in the lv_check_not_in_use fn that gets called before
we do any LV removal/deactivation. In addition to original logic,
this patch adds its own retry loop with a delay (25x0.2 seconds)
besides the existing ioctl retry loop.
Split image should have an out-of-sync attr ('I') - always. Even if
the RAID LV has not been written to since the LV was split off, it is
still not part of the group that makes up the RAID and is therefore
"out-of-sync".
Since the virtual snapshot has no reason to stay alive once we
detach related snapshot - deactivate whole thing in front of
snapshot removal - otherwice the code would get tricky for
support in cluster.
The correct full solution would require to have transactions
for libdm operations.
Also enable to the check for snapshot being opened prior
the origin deactivation, otherwise we could easily end
with the origin being deactivate, but snapshot still kept
active, desynchronizing locking state in cluster.
Recognize DM_SUBSYSTEM_UDEV_FLAG0 which for LVM is the "LVM_NOSCAN"
flag that causes the scanning to be skipped (mainly blkid) and
also directs all the foreign rules to be skipped as well.
Important thing here is that the "watch" udev rules is still set
as well as the /dev/disk/by-id content created (which does not
require any scanning to be done). Also, the flag is dropped on
any subsequent event and scanning done...
A common scenario is during new LV creation when we need to wipe the
newly created LV and avoid any udev scanning before this stage otherwise
it could cause the device (the LV) to be claimed by some other subsystem
for which there were stale metadata within LV data.
This patch adds possibility to mark the LV we're just about to wipe with
a flag that gets passed to udev via DM_COOKIE as a subsystem specific
flag - DM_SUBSYSTEM_UDEV_FLAG0 (in this case the subsystem is "LVM")
so LVM udev rules will take care of handling that.
Patch 562ad293fd introduced code regression
when LV was converted to a thin LV with external origin and at the same time,
conversion of LV to a thin pool has been requested.
(RHBZ: #997704)
data_lv needs to be assigned after test for external conversion find pool.
Accept --ignoreskippedcluster with pvs, vgs, lvs, pvdisplay, vgdisplay,
lvdisplay, vgchange and lvchange to avoid the 'Skipping clustered
VG' errors when requesting information about a clustered VG
without using clustered locking and still exit with success.
The messages can still be seen with -v.
lib/metadata/lv_manip.c:_sufficient_pes_free() was calculating the
required space for RAID allocations incorrectly due to double
accounting. This resulted in failure to allocate when available
space was tight.
When RAID data and metadata areas are allocated together, the total
amount is stored in ah->new_extents and ah->alloc_and_split_meta is
set. '_sufficient_pes_free' was adding the necessary metadata extents
to ah->new_extents without ever checking ah->alloc_and_split_meta.
This often led to double accounting of the metadata extents. This
patch checks 'ah->alloc_and_split_meta' to perform proper calculations
for RAID.
This error is only present in the function that checks for the needed
space, not in the functions that do the actual allocation.
Add allocation/thin_pool_chunk_size_calculation lvm.conf
option to select a method for calculating thin pool chunk
sizes and define two possible values - "default" and "performance".
A previous commit (b6bfddcd0a) which
was designed to prevent segfaults during lvextend when trying to
extend striped logical volumes forgot to include calculations for
RAID4/5/6 parity devices. This was causing the 'contiguous' and
'cling_by_tags' allocation policies to fail for RAID 4/5/6.
The solution is to remember that while we can compare
ah->area_count == prev_lvseg->area_count
for non-RAID, we should compare
(ah->area_count + ah->parity_count) == prev_lvseg->area_count
for a general solution.
lvmetad is not yet supported in clustered environment so
disable it automatically if using lvmconf --enable-cluster
and reset it to default value if using lvmconf --disable-cluster.
Also, add a few comments in lvm.conf about locking_type vs. use_lvmetad
if setting it for clustered environment.
The corosync cluster interface for clvmd did not correctly
deal with node up/down events so that when a node was removed
from the cluster clvmd would prevent remote operations
from happening, as it thought the node was up but not
running clvmd.
This patch fixes that code by simplifying the case to node
being up or down - which was the original intention
and is supported by pacemaker and CPG in the higher layers.
Signed-off-by: Christine Caulfield <ccaulfie@redhat.com>
When NULL info struct is passed in - function is usable
as a quick query for lv_is_active_locally() - with a bonus
we may query for layered device.
So it could be seen as a more efficient lv_is_active_locally().
Add internal devtypes reporting command to display built-in recognised
block device types. (The output does not include any additional
types added by a configuration file.)
> lvm devtypes -o help
Device Types Fields
-------------------
devtype_all - All fields in this section.
devtype_name - Name of Device Type exactly as it appears in /proc/devices.
devtype_max_partitions - Maximum number of partitions. (How many device minor numbers get reserved for each device.)
devtype_description - Description of Device Type.
> lvm devtypes
DevType MaxParts Description
aoe 16 ATA over Ethernet
ataraid 16 ATA Raid
bcache 1 bcache block device cache
blkext 1 Extended device partitions
...
The lvm2-activation-net.service was ordered only with respect to iscsi
and fcoe service before. In addition to that, we also need ordering
with respect to lvm2-activation.service to prevent parallel vgchange -aay
runs which may cause some problems during activation.
See also https://bugs.gentoo.org/show_bug.cgi?id=480066.
With this patch, the ordering is firmly set to:
lvm2-activation-early.service -> lvm2-activation.service -> lvm2-activation-net.service
Thanks to Alexander Tsoy for the original patch (modified a bit here):
https://www.redhat.com/archives/lvm-devel/2013-September/msg00049.html
Add more 'realistic' simulation of dlm locking.
Previous version was not capable to maintain multiple locks.
Current version doesn't handle multiqueues for locks,
so the ordering is different.
The same corner cases that exist for snapshots on mirrors exist for
any logical volume layered on top of mirror. (One example is when
a mirror image fails and a non-repair LVM command is the first to
detect it via label reading. In this case, the LVM command will hang
and prevent the necessary LVM repair command from running.) When
a better alternative exists, it makes no sense to allow a new target
to stack on mirrors as a new feature. Since, RAID is now capable of
running EX in a cluster and thin is not active-active aware, it makes
sense to pair these two rather than mirror+thinpool.
As further background, here are some additional comments that I made
when addressing a bug related to mirror+thinpool:
(https://bugzilla.redhat.com/show_bug.cgi?id=919604#c9)
I am going to disallow thin* on top of mirror logical volumes.
Users will have to use the "raid1" segment type if they want this.
This bug has come down to a choice between:
1) Disallowing thin-LVs from being used as PVs.
2) Disallowing thinpools on top of mirrors.
The problem is that the code in dev_manager.c:device_is_usable() is unable
to tell whether there is a mirror device lower in the stack from the device
being checked. Pretty much anything layered on top of a mirror will suffer
from this problem. (Snapshots are a good example of this; and option #1
above has been chosen to deal with them. This can also be seen in
dev_manager.c:device_is_usable().) When a mirror failure occurs, the
kernel blocks all I/O to it. If there is an LVM command that comes along
to do the repair (or a different operation that requires label reading), it
would normally avoid the mirror when it sees that it is blocked. However,
if there is a snapshot or a thin-LV that is on a mirror, the above code
will not detect the mirror underneath and will issue label reading I/O.
This causes the command to hang.
Choosing #1 would mean that thin-LVs could never be used as PVs - even if
they are stacked on something other than mirrors.
Choosing #2 means that thinpools can never be placed on mirrors. This is
probably better than we think, since it is preferred that people use the
"raid1" segment type in the first place. However, RAID* cannot currently
be used in a cluster volume group - even in EX-only mode. Thus, a complete
solution for option #2 must include the ability to activate RAID logical
volumes (and perform RAID operations) in a cluster volume group. I've
already begun working on this.
Creation, deletion, [de]activation, repair, conversion, scrubbing
and changing operations are all now available for RAID LVs in a
cluster - provided that they are activated exclusively.
The code has been changed to ensure that no LV or sub-LV activation
is attempted cluster-wide. This includes the often overlooked
operations of activating metadata areas for the brief time it takes
to clear them. Additionally, some 'resume_lv' operations were
replaced with 'activate_lv_excl_local' when sub-LVs were promoted
to top-level LVs for removal, clearing or extraction. This was
necessary because it forces the appropriate renaming actions the
occur via resume in the single-machine case, but won't happen in
a cluster due to the necessity of acquiring a lock first.
The *raid* tests have been updated to allow testing in a cluster.
For the most part, this meant creating devices with '-aey' if they
were to be converted to RAID. (RAID requires the converting LV to
be EX because it is a condition of activation for the RAID LV in
a cluster.)
Properly skip unmonitoring of thin pool volume in deactivation code
path. Code makes sure if there is just any thin pool user
it stays monitored with all its resources.
When the pool is created from non-linear target the more complex rules
have to be used and stacking needs to properly decode args for _tdata
LV. Also proper allocation policies are being used according to those
set in lvm2 metadata for data and metadata LVs.
Also properly check for active pool and extra code to active it
temporarily.
With this fix it's now possible to use:
lvcreate -L20 -m2 -n pool vg --alloc anywhere
lvcreate -L10 -m2 -n poolm vg --alloc anywhere
lvconvert --thinpool vg/pool --poolmetadata vg/poolm
lvresize -L+10 vg/pool
Udev daemon has recently introduced a limit on the number of udev
processes (there was no limit before). This causes a problem
when calling pvscan --cache -aay in lvmetad udev rules which
is supposed to activate the volumes. This activation is itself
synced with udev and so it waits for the activation to complete
before the pvscan finishes. The event processing can't continue
until this pvscan call is finished.
But if we're at the limit with the udev process count, we can't
instatiate any more udev processes, all such events are queued
and so we can't process the lvm activation event for which the
pvscan is waiting.
Then we're in a deadlock since the udev process with the
pvscan --cache -aay call waits for the lvm activation udev
processing to complete, but that will never happen as there's
this limit hit with the number of udev processes.
The process with pvscan --cache -aay actually times out eventually
(3min or 30sec, depends on the version of udev).
This patch makes it possible to run the pvscan --cache -aay
in the background so the udev processing can continue and hence
we can avoid the deadlock mentioned above.
The commit 82d83a01ce
"autoactivation: refresh existing VG before autoactivation"
causes problems (dangling udev_sync cookies, slow processing
of the pvscan --cache --major --minor call from udev rules)
when the autoactivation handler is run in parallel on
several PVs that belong to the same VG. Revert this patch
until the exact source of the problem is found and then
properly fixed and handled.
Remove default "/tmp" as destination directory if no args
specified for lvm2-activation-generator. Require all the
args to be specified directly for proper functionality.
The patch allows the user to also pvmove snapshots and origin logical
volumes. This means pvmove should be able to move all segment types.
I have, however, disallowed moving converting or merging logical volumes.
Top-level LVs (like RAID, mirror or thin) are ignored when determining which
portions of an LV to pvmove. If the user specified the name of an LV to
move and it was one of the above types, it would be skipped. The code would
never move on to check whether its sub-LVs needed moving because their names
did not match what the user specified.
The solution is to check whether a sub-LVs is part of the LV whose name was
specified by the user - not just if there was a name match.
This patch allows pvmove to operate on RAID, mirror and thin LVs.
The key component is the ability to avoid moving a RAID or mirror
sub-LV onto a PV that already has another RAID sub-LV on it.
(e.g. Avoid placing both images of a RAID1 LV on the same PV.)
Top-level LVs are processed to determine which PVs to avoid for
the sake of redundancy, while bottom-level LVs are processed
to determine which segments/extents to move.
This approach does have some drawbacks. By eliminating whole PVs
from the allocation list, we might miss the opportunity to perform
pvmove in some senarios. For example, if we have 3 devices and
a linear uses half of the first, a RAID1 uses half of the first and
half of the second, and a linear uses half of the third (FIGURE 1);
we should be able to pvmove the first device (FIGURE 2).
FIGURE 1:
[ linear ] [ -RAID- ] [ linear ]
[ -RAID- ] [ ] [ ]
FIGURE 2:
[ moved ] [ -RAID- ] [ linear ]
[ moved ] [ linear ] [ -RAID- ]
However, the approach we are using would eliminate the second
device from consideration and would leave us with too little space
for allocation. In these situations, the user does have the ability
to specify LVs and move them one at a time.
Do not print success status for lvm2-activation-generator:
"LVM: Activation generator successfully completed."
"LVM: Logical Volume autoactivation enabled." (if use_lvmetad=1)
Though this information is quite useful during boot, it may
be confusing for users if it happens anytime later and it
actually happens if systemd reloads. This is usually on package
update to update the systemd state and load any new units that are
newly installed in the system. The systemd reload is global and
so any existing generators are rerun at that moment too.
This is a regression caused by commit 3bd9048854.
The error message added with that commit "mpath major %d is not dm major %d" is
superfluous.
When scanning for mpath components, we're looking for a parent device.
But this parent device is not necessarily an mpath device (so the dm device)
if it exists - it can be any other device layered on top (e.g. an MD RAID device).
The bug addressed by this patch manifested itself during testing
by showing a mirror that never became 'in-sync' after creation.
The bug is isolated to distributions that do not have support
for openAIS checkpointing (i.e. > RHEL6, > F16).
When a node joins a group that is managing a mirror log, the other
machines in the group send it a checkpoint representing the current
state of the bitmap. More than one machine can send a checkpoint,
but only the initial one should be imported. Once the bitmap state
has been imported from the initial checkpoint, operations (such
as resync, mark, and clear operations) can begin. When subsequent
checkpoints are allowed to be imported, it has the effect of erasing
all the log operations between the initial checkpoint and the ones
that follow.
When cmirrord was updated to handle the absence of openAIS
checkpointing (commit 62e38da133),
the new import_checkpoint() function failed to honor the 'no_read'
parameter. This parameter was designed to avoid reading all but
the initial checkpoint. Honoring this parameter has solved the
issue of corrupting bitmap data with secondary checkpoints.
If loop device is first configured on systems where /dev/loop-control
is used to dynamically create the loop device itself, there's an
ADD+CHANGE even generated. But next time the existing /dev/loop[0-9]*
is reused, there's only a CHANGE event since the device representing
it is already present in kernel (so no ADD event in this case).
We can't ignore this CHANGE event for loop devices! This is a regression
caused by 756bcabbfe. We already had
a similar problem with MD devices which was fixed by
2ac217d408 (but that one was
only an intra-release fix).
When autoactivating a VG, there could be an existing VG with exactly
the same PV UUIDs. The PVs could be reappeared after previous
loss/disconnect (for example disconnecting and reconnecting iscsi).
Since there's no "autodeactivation" yet, the mappings for the LVs
from the VG were left in the system even if the device was disconnected.
These mappings also hold the major:minor of the underlying device.
So if the device reappears, it is assigned a different major:minor
pair (...and kernel name). We need to cope with this during
autoactivation so any existing mappings are corrected for any changes.
The VG refresh does that (the vgchange --refresh functionality) -
call this before VG autoactivation.
(If the VG does not exist yet, the VG refresh is NOP)
Split out the partitioned device filter that needs to open the device
and move the multipath filter in front of it.
When a device is multipathed, sending I/O to the underlying paths may
cause problems, the most obvious being I/O errors visible to lvm if a
path is down.
Revert the incorrect <backtrace> messages added when a device doesn't
pass a filter.
Log each filter initialisation to show sequence.
Avoid duplicate 'Using $device' debug messages.
Recent version of util-linux/umount (v2.23+) provides
umount --all-targets that can unmount all the mount targets of
the same device (the bind mounts). Use this if available when
calling the umount blkdeactivate.
Otherwise, for older versions of util-linux, use findmnt
(that is also a part of the util-linux) to iterate over all
mount targets of the same device - this is the manual way.
The blkdeactivate now suppresses error messages from external
tools that are called. Instead, only a summary message "done"
or "skipped" is issued by blkdeactivate as any error in calling
the external tool (e.g. unmounting or deactivating a device) causes
the device to be skipped and the blkdeactivate continues with the
next device in the tree.
Add new -e/--errors switch to display any error messages from
external tools.
Also, suppress any output given by the external tools and add
new -v/--verbose switch to display it including the verbose
output of the tools called (this will enable error reporting
as well).
Also add blkdeactivate -vv for even more debug (the script's debug).
84 files changed, 1540 insertions(+), 442 deletions(-)
Mostly bug fixes this time.
Also note:
md raid replaces dm mirroring as the default implementation.
Can call out to thin_repair to fix thin metadata.
Improved clvmd error detection/debugging information.
According to bug 995193, if a volume group
1) contains a mirror
2) is clustered
3) 'locking_type' = 0 is used
then it is not possible to remove the 'c'luster flag from the VG. This
is due to the way _lv_is_active behaves.
We shouldn't allow the cluster flag to be flipped unless the mirrors in
the cluster are not active. This is because different kernel modules
are used depending on whether a mirror is cluster or not. When we
attempt to see if the mirror is active, we first check locally. If it
is not, then we attempt to check for remotely active instances if the VG
is clustered. Since the no_lock locking type is LCK_CLUSTERED, but does
not implement 'query_resource', remote_lock_held will always return an
error in this case. An error from remove_lock_held is treated as though
the lock _is_ held (i.e. the LV is active remotely). This blocks the
cluster flag from changing.
The solution is to implement 'query_resource' for the no_lock type. It
will report a message and return 1. This will allow _lv_is_active to
function properly. The LV would be considered not active remotely and
the VG can change its flag.
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.
