]> sourceware.org Git - lvm2.git/log
lvm2.git
7 years agotests: aux.sh drop unnecessary $
Zdenek Kabelac [Wed, 28 Jun 2017 22:17:59 +0000 (00:17 +0200)]
tests: aux.sh drop unnecessary $

$/${} is unnecessary on arithmetic variables.

7 years agotests: aux.sh use pgrep
Zdenek Kabelac [Wed, 28 Jun 2017 22:04:47 +0000 (00:04 +0200)]
tests: aux.sh use pgrep

7 years agotests: aux.sh double quote
Zdenek Kabelac [Wed, 28 Jun 2017 22:02:19 +0000 (00:02 +0200)]
tests: aux.sh double quote

7 years agoclvmd_fix_conf.sh: double quote
Zdenek Kabelac [Thu, 29 Jun 2017 20:07:49 +0000 (22:07 +0200)]
clvmd_fix_conf.sh: double quote

7 years agoclvmd_fix_conf.sh: legacy code
Zdenek Kabelac [Thu, 29 Jun 2017 20:03:00 +0000 (22:03 +0200)]
clvmd_fix_conf.sh: legacy code

Use $(..) instead of legacy `..`.

7 years agovgimportclone.sh: do not quote regex
Zdenek Kabelac [Thu, 29 Jun 2017 16:54:26 +0000 (18:54 +0200)]
vgimportclone.sh: do not quote regex

Don't quote rhs of =~, it'll match literally rather than as a regex.

7 years agovgimportclone.sh: use well defined tests
Zdenek Kabelac [Wed, 28 Jun 2017 18:35:34 +0000 (20:35 +0200)]
vgimportclone.sh: use well defined tests

Prefer [ p ] && [ q ] as [ p -a q ] is not well defined.
Prefer [ p ] || [ q ] as [ p -o q ] is not well defined.

7 years agovgimportclone.sh: drop extra $
Zdenek Kabelac [Wed, 28 Jun 2017 18:31:09 +0000 (20:31 +0200)]
vgimportclone.sh: drop extra $

$/${} is unnecessary on arithmetic variables.

7 years agovgimportclone.sh: add -r to read
Zdenek Kabelac [Wed, 28 Jun 2017 18:26:23 +0000 (20:26 +0200)]
vgimportclone.sh: add -r to read

read without -r will mangle backslashes.

7 years agovgimportclone.sh: avoid legacy syntax
Zdenek Kabelac [Wed, 28 Jun 2017 18:20:15 +0000 (20:20 +0200)]
vgimportclone.sh: avoid legacy syntax

Use $(..) instead of legacy `..`.

7 years agovgimportclone.sh: cannot trap signal 9
Zdenek Kabelac [Wed, 28 Jun 2017 18:06:59 +0000 (20:06 +0200)]
vgimportclone.sh: cannot trap signal 9

7 years agovgimportclone.sh: double quote
Zdenek Kabelac [Wed, 28 Jun 2017 18:05:42 +0000 (20:05 +0200)]
vgimportclone.sh: double quote

7 years agovgimportclone.sh: no mixing string and array
Zdenek Kabelac [Wed, 28 Jun 2017 17:40:34 +0000 (19:40 +0200)]
vgimportclone.sh: no mixing string and array

Argument mixes string and array. Use * or separate argument.

7 years agolvm2_cluster_activation_red_hat.sh: double quote
Zdenek Kabelac [Thu, 29 Jun 2017 20:12:23 +0000 (22:12 +0200)]
lvm2_cluster_activation_red_hat.sh: double quote

7 years agolvmconf_lockingtype2: double quote
Zdenek Kabelac [Wed, 28 Jun 2017 19:13:52 +0000 (21:13 +0200)]
lvmconf_lockingtype2: double quote

7 years agolvmconf_lockingtype2: preferable syntax
Zdenek Kabelac [Wed, 28 Jun 2017 19:09:52 +0000 (21:09 +0200)]
lvmconf_lockingtype2: preferable syntax

Prefer [ p ] || [ q ] as [ p -o q ] is not well defined.

7 years agolvmdump: replace individual redirects
Zdenek Kabelac [Thu, 29 Jun 2017 14:21:19 +0000 (16:21 +0200)]
lvmdump: replace individual redirects

Consider using { cmd1; cmd2; } >> file instead of individual redirects.

7 years agolvmdump: drop useless cat
Zdenek Kabelac [Wed, 28 Jun 2017 19:02:23 +0000 (21:02 +0200)]
lvmdump: drop useless cat

Useless cat. Consider 'cmd < file | ..' or 'cmd file | ..' instead.

7 years agolvmdump: legacy syntax
Zdenek Kabelac [Wed, 28 Jun 2017 19:00:48 +0000 (21:00 +0200)]
lvmdump: legacy syntax

Use $(..) instead of legacy `..`.

7 years agolvmdump: drop extra $
Zdenek Kabelac [Wed, 28 Jun 2017 18:56:17 +0000 (20:56 +0200)]
lvmdump: drop extra $

$/${} is unnecessary on arithmetic variables.

7 years agolvmdump: replace printf
Zdenek Kabelac [Wed, 28 Jun 2017 18:54:00 +0000 (20:54 +0200)]
lvmdump: replace printf

Don't use variables in the printf format string. Use printf "..%s.." "$foo".
Advice replaced with plain 'echo'.

7 years agolvmdump: double quote
Zdenek Kabelac [Wed, 28 Jun 2017 18:48:19 +0000 (20:48 +0200)]
lvmdump: double quote

Double quote to prevent globbing and word splitting.

7 years agolvmconf: prefer well defined and
Zdenek Kabelac [Wed, 28 Jun 2017 17:25:41 +0000 (19:25 +0200)]
lvmconf: prefer well defined and

Prefer [ p ] && [ q ] as [ p -a q ] is not well defined.

7 years agolvmconf: double quote
Zdenek Kabelac [Wed, 28 Jun 2017 15:22:08 +0000 (17:22 +0200)]
lvmconf: double quote

7 years agolvmconf: fix test for setting external locking type
Zdenek Kabelac [Wed, 28 Jun 2017 15:22:22 +0000 (17:22 +0200)]
lvmconf: fix test for setting external locking type

Fix test for check of set locking lib dir.

7 years agoblkdeactivate: quote literal
Zdenek Kabelac [Thu, 29 Jun 2017 16:47:28 +0000 (18:47 +0200)]
blkdeactivate: quote literal

This { is literal. Check expression (missing ;/\n?) or quote it.

7 years agoblkdeactivate: more readable if-then-else-fi
Zdenek Kabelac [Wed, 28 Jun 2017 14:55:38 +0000 (16:55 +0200)]
blkdeactivate: more readable if-then-else-fi

7 years agoblkdeactivate: split local declaration and assign
Zdenek Kabelac [Wed, 28 Jun 2017 14:50:34 +0000 (16:50 +0200)]
blkdeactivate: split local declaration and assign

7 years agoblkdeactivate: drop unused var
Zdenek Kabelac [Wed, 28 Jun 2017 17:36:49 +0000 (19:36 +0200)]
blkdeactivate: drop unused var

7 years agoblkdeactivate: double quote
Zdenek Kabelac [Wed, 28 Jun 2017 14:33:00 +0000 (16:33 +0200)]
blkdeactivate: double quote

Using double quote to prevent globbing and word splitting.

7 years agoblkdeactivate: replace obsoleted syntax
Zdenek Kabelac [Wed, 28 Jun 2017 14:15:36 +0000 (16:15 +0200)]
blkdeactivate: replace obsoleted syntax

7 years agofsadm: make if statement more clear
Zdenek Kabelac [Wed, 28 Jun 2017 14:13:29 +0000 (16:13 +0200)]
fsadm: make if statement more clear

7 years agofsadm: add -r to read
Zdenek Kabelac [Wed, 28 Jun 2017 14:07:33 +0000 (16:07 +0200)]
fsadm: add -r to read

read without -r will mangle backslashes

7 years agofsadm: no mixing string and array
Zdenek Kabelac [Thu, 29 Jun 2017 14:01:42 +0000 (16:01 +0200)]
fsadm: no mixing string and array

Argument mixes string and array. Use * or separate argument.

7 years agofsadm: split local var dev
Zdenek Kabelac [Wed, 28 Jun 2017 14:02:13 +0000 (16:02 +0200)]
fsadm: split local var dev

Split defining of local variables from being them
assigned at the same time (lossing error code).

7 years agofsadm: remove surrounding $()
Zdenek Kabelac [Thu, 29 Jun 2017 14:04:19 +0000 (16:04 +0200)]
fsadm: remove surrounding $()

Remove surrounding $() to avoid executing output

7 years agofsadm: fix shell globbing
Zdenek Kabelac [Thu, 29 Jun 2017 11:15:47 +0000 (13:15 +0200)]
fsadm: fix shell globbing

Fixing reported issue for shellcheck.
Use "" properly to avoid having shell variables split.

7 years agopost-release
Alasdair G Kergon [Wed, 28 Jun 2017 16:00:49 +0000 (17:00 +0100)]
post-release

7 years agopre-release v2_02_172
Alasdair G Kergon [Wed, 28 Jun 2017 15:38:46 +0000 (16:38 +0100)]
pre-release

7 years agocommand: we don't have all standard macros
Zdenek Kabelac [Wed, 28 Jun 2017 13:05:16 +0000 (15:05 +0200)]
command: we don't have all standard macros

So just quite warning

7 years agocoverity: checked_return of close
Zdenek Kabelac [Wed, 28 Jun 2017 12:40:35 +0000 (14:40 +0200)]
coverity: checked_return of close

Check (or make quiet) close() ret code.
NOTE: there is another duplicated code of daemonize function which
should be converted to libdaemon.

7 years agoman: regenerate
Alasdair G Kergon [Tue, 27 Jun 2017 14:47:00 +0000 (15:47 +0100)]
man: regenerate

7 years agolog: fix message prefix truncation
Alasdair G Kergon [Tue, 27 Jun 2017 11:41:59 +0000 (12:41 +0100)]
log: fix message prefix truncation

Restore the 29-character prefix limit that commit
e907535ab977ea29319aeb015cec6c06729f7ce4 ("coverity: use better
function") accidentally reduced to 28 characters.

7 years agocoverity: model few more functions
Zdenek Kabelac [Tue, 27 Jun 2017 07:23:22 +0000 (09:23 +0200)]
coverity: model few more functions

Reduce amount of FP.

7 years agocoverity: move initilization of count variable
Zdenek Kabelac [Tue, 27 Jun 2017 10:05:27 +0000 (12:05 +0200)]
coverity: move initilization of count variable

Make code more understandble for Coverity, so it sees connection
between 'extents' and 'count' and in fact code is more readable.

7 years agocoverity: drop unneeded assigns
Zdenek Kabelac [Tue, 27 Jun 2017 09:38:56 +0000 (11:38 +0200)]
coverity: drop unneeded assigns

7 years agocoverity: validate length of renaming path
Zdenek Kabelac [Tue, 27 Jun 2017 09:37:24 +0000 (11:37 +0200)]
coverity: validate length of renaming path

Make sure path fits into buffer on stack.

7 years agocoverity: check malloc results
Zdenek Kabelac [Tue, 27 Jun 2017 08:18:00 +0000 (10:18 +0200)]
coverity: check malloc results

7 years agocoverity: use better function
Zdenek Kabelac [Tue, 27 Jun 2017 08:17:33 +0000 (10:17 +0200)]
coverity: use better function

7 years agocoverity: extra check for find_pool_seg
Zdenek Kabelac [Tue, 27 Jun 2017 08:16:13 +0000 (10:16 +0200)]
coverity: extra check for find_pool_seg

find_pool_seg may return NULL in some internal error stats.
Handle it explicitely.

7 years agodebug: more display_lvname usage
Zdenek Kabelac [Tue, 27 Jun 2017 06:28:36 +0000 (08:28 +0200)]
debug: more display_lvname usage

7 years agocoverity: validate retval from get_only_segment_using_this_lv
Zdenek Kabelac [Mon, 26 Jun 2017 22:24:34 +0000 (00:24 +0200)]
coverity: validate retval from get_only_segment_using_this_lv

7 years agocoverity: add some error path for failed allocs
Zdenek Kabelac [Mon, 26 Jun 2017 21:57:12 +0000 (23:57 +0200)]
coverity: add some error path for failed allocs

Coverity reports some unchecked allocations.

7 years agodebug: use display_lvname
Zdenek Kabelac [Mon, 26 Jun 2017 21:58:06 +0000 (23:58 +0200)]
debug: use display_lvname

7 years agodebug: fix error path
Zdenek Kabelac [Mon, 26 Jun 2017 21:32:04 +0000 (23:32 +0200)]
debug: fix error path

Coverity:
log_error may have dereference NULL pointer - use directly
segtype string name.

7 years agodebug: fail in backup is not traced nor error
Zdenek Kabelac [Mon, 26 Jun 2017 21:27:10 +0000 (23:27 +0200)]
debug: fail in backup is not traced nor error

7 years agodebug: tracing error path
Zdenek Kabelac [Mon, 26 Jun 2017 21:21:22 +0000 (23:21 +0200)]
debug: tracing error path

7 years agodebug: check for fail in id validation
Zdenek Kabelac [Mon, 26 Jun 2017 21:20:29 +0000 (23:20 +0200)]
debug: check for fail in id validation

7 years agocleanup: remove unused code
Zdenek Kabelac [Mon, 26 Jun 2017 21:19:57 +0000 (23:19 +0200)]
cleanup: remove unused code

7 years agodev_manager: add missing validation for usable target_name
Zdenek Kabelac [Mon, 26 Jun 2017 21:34:38 +0000 (23:34 +0200)]
dev_manager: add missing validation for usable target_name

Coverity shown path where target_name == NULL would be dereferenced.

7 years agoraid: fix write_commit_backup
Zdenek Kabelac [Mon, 26 Jun 2017 21:25:24 +0000 (23:25 +0200)]
raid: fix write_commit_backup

With commit 41c10034aa0db3c63f18ff78b64af3a799c154a1 we actually
do require LV to be used with _vg_write_lv_suspend_commit_backup().

So write a proper separte single wrapper for write && commit && backup.

7 years agocmdline: fix missing NULL at end of array of args
Zdenek Kabelac [Mon, 26 Jun 2017 18:03:05 +0000 (20:03 +0200)]
cmdline: fix missing NULL at end of array of args

lvm_run needs to place NULL as the last element into argv[].
Otherwise we get:

Conditional jump or move depends on uninitialised value(s)
 _command_required_pos_matches (lvmcmdline.c:1443)
 _find_command (lvmcmdline.c:1610)
 lvm_run_command (lvmcmdline.c:2770)
 lvm2_run (lvmcmdlib.c:91)

7 years agolibdm: fix initialization of head for reused structure
Zdenek Kabelac [Mon, 26 Jun 2017 16:22:40 +0000 (18:22 +0200)]
libdm: fix initialization of head for reused structure

Dmeventd reuses  'dm_task' struct for some STATUS operation, but due to
missing reinitization of dm_task target list, it has caused misprocesing
of recieved events as the parsed target has been simply added to the
list of existing status and cause multiple actions being called for
single event.

7 years agotests: show more debug info
Zdenek Kabelac [Sat, 24 Jun 2017 21:01:26 +0000 (23:01 +0200)]
tests: show more debug info

Helpful for analysis (revert later).

7 years agotests: slightly longer delay
Zdenek Kabelac [Sat, 24 Jun 2017 20:58:08 +0000 (22:58 +0200)]
tests: slightly longer delay

Try to avoid race with slightly longer delay.

7 years agoraid: allow more sync action for extraction
Zdenek Kabelac [Sat, 24 Jun 2017 20:28:25 +0000 (22:28 +0200)]
raid: allow more sync action for extraction

Since we discovered status reporting from 'md' goes from large set
of weird states we can't just decided based on this word.
So let it pass for rebuild and idle as well
and check for health devices afterwards.

7 years agotests: correct percent values are reported
Zdenek Kabelac [Sat, 24 Jun 2017 15:24:38 +0000 (17:24 +0200)]
tests: correct percent values are reported

7 years agotests: unit test for percent printing
Zdenek Kabelac [Sat, 24 Jun 2017 11:33:36 +0000 (13:33 +0200)]
tests: unit test for percent printing

7 years agocleanup: use display_percent
Zdenek Kabelac [Sat, 24 Jun 2017 14:22:36 +0000 (16:22 +0200)]
cleanup: use display_percent

Replace occurence of %.2f with call of display_percent function.

7 years agodisplay: add display_percent function
Zdenek Kabelac [Sat, 24 Jun 2017 14:03:21 +0000 (16:03 +0200)]
display: add display_percent function

Add universal function to print dm_percent_t values via single
code path.

TODO: extend with configurale precision of printed values.

7 years agolibdm: use rounded float for percent print
Zdenek Kabelac [Sat, 24 Jun 2017 14:39:50 +0000 (16:39 +0200)]
libdm: use rounded float for percent print

Use new added  dm_percent_to_round_float to enhance print
of percentage values.

7 years agolibdm: implement dm_percent_to_round_float
Zdenek Kabelac [Sat, 24 Jun 2017 11:24:26 +0000 (13:24 +0200)]
libdm: implement dm_percent_to_round_float

Add function to adjust printing of percent values in better way.
Rounding here is going along following rules:

0% & 100% are always clearly reported with  .0 decimal points.

Values slightly above 0% we make sure a nearest bigger
non zero value with given precission is printed
(i.e. 0.01  for  %.2f  will be shown)

For values closely approaching 100% we again detect and adjust value
that is less then 100 when printed.
(i.e. 99.99  for %.2f will be shown).

For other values we are leaving them with standard rounding mechanism
since we care mainly about corner values 0 & 100 which need to be
printed precisely.

7 years agodmeventd: improve more raid status reporting
Zdenek Kabelac [Fri, 23 Jun 2017 22:06:12 +0000 (00:06 +0200)]
dmeventd: improve more raid status reporting

When we want to report primary leg failure, check for intial 'a',
since otherwice 'Aa idle' is normally visible.

Also reset array of bit flags marking dead devices, once
plugin detects raid is in sync.

7 years agotests: use allocate policy in raid test
Zdenek Kabelac [Fri, 23 Jun 2017 21:22:34 +0000 (23:22 +0200)]
tests: use allocate policy in raid test

7 years agotests: add transient failure test
Zdenek Kabelac [Fri, 23 Jun 2017 21:20:01 +0000 (23:20 +0200)]
tests: add transient failure test

Check raid  repairs still 'present', but failed device.

TODO: tests needs more checking about repair actually doing its work.

7 years agotests: require also no 'a' for in-sync
Zdenek Kabelac [Fri, 23 Jun 2017 19:14:29 +0000 (21:14 +0200)]
tests: require also no 'a' for in-sync

Raid easily reports  'idle' with 'a', so in this case
we will not report in-sync.

7 years agoraid: plugin does not to use --config
Zdenek Kabelac [Fri, 23 Jun 2017 19:15:37 +0000 (21:15 +0200)]
raid: plugin does not to use --config

Functionality of ignore suspend devices is already granted by:

lvm2_disable_dmeventd_monitoring() -> init_run_by_dmeventd() ->
init_ignore_suspended_devices().

In fact plugins should never use --config because it has
some unpleasant technical issues.

7 years agoraid: recognize transient failed raid leg
Zdenek Kabelac [Fri, 23 Jun 2017 20:50:19 +0000 (22:50 +0200)]
raid: recognize transient failed raid leg

When raid leg rimage device is marked as 'D'ead by mdcore,
lvm2 was not able to replace such device with allocate policy,
as device has not appared as missing.

Add detection of transiently failing devices.

7 years agotests: request 1.12 for reshaping raid
Zdenek Kabelac [Fri, 23 Jun 2017 12:52:34 +0000 (14:52 +0200)]
tests: request 1.12 for reshaping raid

We will need  1.12 as the kernel needs some mandatory md fixes,
otherwise random freezes may appear.

7 years agotests: handle su login difference
Zdenek Kabelac [Fri, 23 Jun 2017 13:11:26 +0000 (15:11 +0200)]
tests: handle su login difference

Workaround case, where it matters if 'su' was user with '-' or without.

7 years agocleanup: update messages
Zdenek Kabelac [Fri, 23 Jun 2017 08:59:12 +0000 (10:59 +0200)]
cleanup: update messages

7 years agodebug: add missing internal error message
Zdenek Kabelac [Fri, 23 Jun 2017 12:41:10 +0000 (14:41 +0200)]
debug: add missing internal error message

Do not just 'return_0'  log error would need to be shown.

7 years agocache: restore origin only reload
Zdenek Kabelac [Fri, 23 Jun 2017 15:05:24 +0000 (17:05 +0200)]
cache: restore origin only reload

Basically reverting commit 58a9f88b8c021a5e056b883053f257f1a898adf7.
We can use origin_only  in case we are snapshot's origin,
as we do support this stack.

So when we are 'uncaching'  origin+snaps - we do need to reload only
origin and we do not need to play with snaps.

7 years agoraid: switch message to verbose
Zdenek Kabelac [Fri, 23 Jun 2017 09:14:31 +0000 (11:14 +0200)]
raid: switch message to verbose

As this is not 'error' resulting query, decrease reported level.

7 years agoraid: improving messages for regionsize change
Zdenek Kabelac [Fri, 23 Jun 2017 12:40:10 +0000 (14:40 +0200)]
raid: improving messages for regionsize change

Handle change of 'region size' better and follow also standard rule
if the command can't success (i.e. size is already same) we return
error for all such cases.

Also log_pring more info about adjusted value (just like we
do for rounding)

Also avoid keep pointers on 'display_*' values - they are in
ringbuffer for immediate use - not to be kept across multiple calls
(as they could be already overwritten by later calls) - so dropped
seg_region_size_str

7 years agotests: flushing of cache is abortable
Zdenek Kabelac [Thu, 22 Jun 2017 18:09:00 +0000 (20:09 +0200)]
tests: flushing of cache is abortable

7 years agotests: check lvdisplay for cache works
Zdenek Kabelac [Thu, 22 Jun 2017 18:08:24 +0000 (20:08 +0200)]
tests: check lvdisplay for cache works

7 years agodebug: show message only when origin_only was set
Zdenek Kabelac [Thu, 22 Jun 2017 15:14:47 +0000 (17:14 +0200)]
debug: show message only when origin_only was set

7 years agocache: fix lvdisplay output
Zdenek Kabelac [Thu, 22 Jun 2017 18:05:26 +0000 (20:05 +0200)]
cache: fix lvdisplay output

Unused cache pool may have lots of fields actually undefined,
so avoid printing them, if they are not specified in metadata.

7 years agocache: fix lvdisplay --maps
Zdenek Kabelac [Thu, 22 Jun 2017 17:53:27 +0000 (19:53 +0200)]
cache: fix lvdisplay --maps

'lvdisplay -m' tried to go through NULL policy settings,
when such policy was not defined for CachedLV.

Patch is fixing display of cache-pool without defined settings,
as this is now a valid pool and we mostly want users to define
these settings when actually really caching a LV.

7 years agocache: drop usage of origin_only
Zdenek Kabelac [Thu, 22 Jun 2017 15:12:27 +0000 (17:12 +0200)]
cache: drop usage of origin_only

Since cache LV can be a stacked device, there is no real reason
trying to use slight optimised tree for origin_only cache reload
(it could be even wrongly implemented in this case).

We can easily go with stardard tree load here.

7 years agocache: make syncing abortable by user
Zdenek Kabelac [Thu, 22 Jun 2017 15:08:47 +0000 (17:08 +0200)]
cache: make syncing abortable by user

When user runs command like 'lvconvert --splitcache' the operation
might be actually either slow or not making any progress in kernel,
so lets give user a chance to abort such operation.

When user press 'Ctrl+C' device table is restored to pre-flushing state.

7 years agomirror: reformat conditional
Heinz Mauelshagen [Wed, 21 Jun 2017 22:57:16 +0000 (00:57 +0200)]
mirror: reformat conditional

7 years agoraid: fix segfault
Heinz Mauelshagen [Wed, 21 Jun 2017 22:49:00 +0000 (00:49 +0200)]
raid: fix segfault

Add missing else clause
(already missing in initial commit fe18e5e77a488618510648c7801ea3375ce5aac4).

Resolves: rhbz1463794

7 years agotests: lvresize needs --yes
Zdenek Kabelac [Wed, 21 Jun 2017 11:58:55 +0000 (13:58 +0200)]
tests: lvresize needs --yes

7 years agolvresize: support passing --yes to fsadm
Zdenek Kabelac [Wed, 21 Jun 2017 12:02:57 +0000 (14:02 +0200)]
lvresize: support passing --yes to fsadm

Since fsadm now needs --yes to pass prompting operations,
we need to pass --yes from  lvresize to fsadm.

7 years agoraid: update path for repair
Zdenek Kabelac [Wed, 21 Jun 2017 11:57:22 +0000 (13:57 +0200)]
raid: update path for repair

Updating path from commit 61980bcf0624790bfdc07d371fd510d1c4079fb5.

When repair is running, no removing PVS are given so it shall return
success in such case.

7 years agodebug: missing traces
Zdenek Kabelac [Wed, 21 Jun 2017 09:29:16 +0000 (11:29 +0200)]
debug: missing traces

7 years agoraid: use log_error on error path
Zdenek Kabelac [Wed, 21 Jun 2017 09:58:38 +0000 (11:58 +0200)]
raid: use log_error on error path

Converting log_warn to log_error since error must be logged
when tool returns error.

7 years agoraid: report error when specified devices are not contained
Zdenek Kabelac [Wed, 21 Jun 2017 09:45:20 +0000 (11:45 +0200)]
raid: report error when specified devices are not contained

lvm2 always return non-zero error code when action cannot happen.

This page took 0.069108 seconds and 5 git commands to generate.