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.
These wrappers have been replaced by direct calls
to vg_read() and find_lv() in previous commits.
This commit should have no functional impact since
all bits were already unreachable.
Previous patch incorrectly skipped replace of @LOCALEDIR@.
The standard option is --localedir so use --with-localedir
as backward compatible option and set localedir if it's not
yet been set (if the could ever happen).
Use double-eval to translate $datarootdir to $prefix to real dir.
Use CFG_DEFAULT_COMMENTED and CFG_DEFAULT_UNDEFINED to
replicate the existing comments in example.conf.
Fix host_list to be cfg_array.
UNDEFINED is only used if the value depends on other
system/kernel values outside of lvm. The most common
case is when dm-thin or dm-cache have built-in default
settings in the kernel, and lvm will use those built-in
default values unless the corresponding lvm config setting
is set.
COMMENTED is used to comment out the default setting in
lvm.conf. The effect is that if the LVM version is
upgraded, and the new version of LVM has new built-in
default values, the new defaults are used by LVM unless
the previous default value was set (uncommented) in lvm.conf.
Introduce new implmentation of dm_task_get_info() function
with support for reading internal_suspend.
.
This time it is done in a 'versioned' way.
We keep the old fashion dm_task_get_info(Base) to implement
the old behavior of 1.02.95 libdm code.
libdm version 1.02.96 introduced 'macro' wrapper
dm_task_get_info_with_deferred_remove with new implementation
of dm_task_get_info() - we cannot do anything else then to
provide compatible version of this symbol.
Now in version 1.02.97 we add new versioned implementation of
dm_task_get_info(DM_1_02_97) symbol.
This has the effect that i.e. rpm build will finaly resolve proper
dependency on a new symbol - so it will be no longer possible,
to build a new binary and use old library
(rpm -q --provides will show libdevmapper.so.1.02(DM_1_02_97)(64bit))
Also the history is now tracked. If a new function is added (or
reimplemented), it needs to be placed in proper file,
so it could be exported with right versioning symbol.
File .exported_symbols.Base should and any existing older DM
should be treated as read-only after a release.
Also - only libdm has been currently enhanced with versioned .Base
file, as soon as other libs (liblvm, libdevmapper-event) needs changes
they should also get their exported symbol files - meanwhile
make.tmpl handles both cases.
Configure provides proper settings for
use_lvmetad and use_lvmpolld conf setttings.
When the build of polld & lvmetad, these settings
are enabled by default unless explicitelly disabled
with --disable-use-lvmetad/--disable-use-lvmpolld.
This is an alternative/equivalent to commit
ca67cf84df
The problem (wrong label->dev after a new preferred
duplicate device is chosen) was isolated to the lvmetad
case (non-lvmetad worked fine), and this fixes the problem
by setting the new label->dev in the lvmetad-specific
code rather than in the general lvmcache code.
In process_each_{vg,lv,pv} when no vgname args are given,
the first step is to get a list of all vgid/vgname on the
system. This is exactly what lvmetad returns from a
vg_list request. The current code is doing a vg_lookup
on each VG after the vg_list and populating lvmcache with
the info for each VG. These preliminary vg_lookup's are
unnecessary, because they will be done again when the
processing functions call vg_read. This patch eliminates
the initial round of vg_lookup's, which can roughly cut in
half the number of lvmetad requests and save a lot of extra work.
Use 64bit arithmentic for PV size calculation (Coverity).
Also remove sector shift for compared PV size, since all
values are already held in sectors.
This fixes validatio of PV size when restoring PV
from vg metadata backup file.
Synopsis: STR_LIST needs to be treated as STR for properties.
For any lvm property that was internally 'typed' as a string list we were failing
to return a string in the property API. This was due to the fact that for the
properties to work the value needs to either be evaulated as a string or a
number. This change corrects the macro used to build the memory array of
structures so that the string bitfield is set as needed to ensure that the value
is a string.
https://bugzilla.redhat.com/show_bug.cgi?id=1139920
Signed-off-by: Tony Asleson <tasleson@redhat.com>
Currently lvm2app properties have the following structure:
typedef struct lvm_property_value {
uint32_t is_settable:1;
uint32_t is_string:1;
uint32_t is_integer:1;
uint32_t is_valid:1;
uint32_t padding:28;
union {
const char *string;
uint64_t integer;
} value;
} lvm_property_value_t;
which assumes that numerical values were in the range of 0 to 2**64-1. However,
some of the properties were 'signed', like LV major/minor numbers and some
reserved values for properties that represent percentages. Thus when the
values were retrieved they were in two's complement notation. So for a -1
major number the API user would get a value of 18446744073709551615. The
API user could cast the returned value to an int64_t to handle this, but that
requires the API developer to look at the source code and determine when it
should be done.
This change modifies the return property structure to:
typedef struct lvm_property_value {
uint32_t is_settable:1;
uint32_t is_string:1;
uint32_t is_integer:1;
uint32_t is_valid:1;
uint32_t is_signed:1;
uint32_t padding:27;
union {
const char *string;
uint64_t integer;
int64_t signed_integer;
} value;
} lvm_property_value_t;
With this addition the API user can interrogate that the value is numerical,
(is_integer = 1) and subsequently check if it's signed (is_signed = 1) too.
If signed, then the API developer should use the union's signed_integer to
avoid casting.
This change maintains backwards compatibility as the structure size remains
unchanged and integer value remains unchanged. Only the additional bit
taken from the pad is utilized.
Bugzilla reference:
https://bugzilla.redhat.com/show_bug.cgi?id=838257
Signed-off-by: Tony Asleson <tasleson@redhat.com>
When kernel target reports sync status as 0% it might as well mean
it's 100% in sync, just the target is in some race inconsistent
state - so reread once again and take a more optimistic value ;)
Patch tries to work around:
https://bugzilla.redhat.com/show_bug.cgi?id=1210637
This patch adds supporting code for handling deprecated settings.
Deprecated settings are not displayed by default in lvmconfig output
(except for --type current and --type diff). There's a new
"--showdeprecated" lvmconfig option to display them if needed.
Also, when using lvmconfig --withcomments, the comments with info
about deprecation are displayed for deprecated settings and with
lvmconfig --withversions, the version in which the setting was
deprecated is displayed in addition to the version of introduction.
If using --atversion with a version that is lower than the one
in which the setting was deprecated, the setting is then considered
as not deprecated (simply because at that version it was not
deprecated).
For example:
$ lvmconfig --type default activation
activation {
...
raid_region_size=512
...
}
$ lvmconfig --type default activation --showdeprecated
activation {
...
mirror_region_size=512
raid_region_size=512
...
}
$ lvmconfig --type default activation --showdeprecated --withversions
activation {
...
# Available since version 1.0.0.
# Deprecated since version 2.2.99.
mirror_region_size=512
# Available since version 2.2.99.
raid_region_size=512
...
}
$ lvmconfig --type default activation --showdeprecated --withcomments
activation {
...
# Configuration option activation/mirror_region_size.
# This has been replaced by the activation/raid_region_size
# setting.
# Size (in KB) of each copy operation when mirroring.
# This configuration option is deprecated.
mirror_region_size=512
# Configuration option activation/raid_region_size.
# Size in KiB of each raid or mirror synchronization region.
# For raid or mirror segment types, this is the amount of
# data that is copied at once when initializing, or moved
# at once by pvmove.
raid_region_size=512
...
}
$ lvmconfig --type default activation --withcomments --atversion 2.2.98
activation {
...
# Configuration option activation/mirror_region_size.
# Size (in KB) of each copy operation when mirroring.
mirror_region_size=512
...
}
A preparatory code for marking configuration nodes as deprecated:
- struct cfg_def_item gains 2 new fields ("deprecated_since_version" and "deprecation_comment"
- cfg* macros to handle new fields
- related config_settings.h edits to add new fields for each item (null for all at the moment)
Patch with implementation will follow...
Before this patch:
$ lvmconfig --type list --withversions --withsummary global/use_lvmetad
global/use_lvmetad - Use lvmetad to cache metadata and reduce disk scanning. [2.2.93]
$ lvmconfig --type list --withversions global/use_lvmetad
global/use_lvmetad
With this patch applied:
$ lvmconfig --type list --withversions --withsummary global/use_lvmetad
global/use_lvmetad - Use lvmetad to cache metadata and reduce disk scanning. [2.2.93]
$ lvmconfig --type list --withversions global/use_lvmetad
global/use_lvmetad - [2.2.93]
We're commenting out settings with undefined default values.
The comment character '#' was printed at the very beginning of
the line, it should be placed just at the beginning of the setting,
after the space/tab prefix is printed.
Before this patch:
$ lvmconfig --type default activation
activation {
...
# volume_list=[]
...
}
With this patch applied:
$ lvmconfig --type default activation
activation {
...
# volume_list=[]
...
}
These settings are in the "unsupported" group:
devices/loopfiles
log/activate_file
metadata/disk_areas (section)
metadata/disk_areas/<disk_area> (section)
metadata/disk_areas/<disk_area>/size
metadata/disk_areas/<disk_area>/id
These settings are in the "advanced" group:
devices/dir
devices/scan
devices/types
global/proc
activation/missing_stripe_filler
activation/mlock_filter
metadata/pvmetadatacopies
metadata/pvmetadataignore
metadata/stripesize
metadata/dirs
Also, this patch causes the --ignoreunsupported and --ignoreadvanced
switches to be honoured for all config types (lvmconfig --type).
By default, the --type current and --type diff display unsupported
settings, the other types ignore them - this patch also introduces
--showunsupported switch for all these other types to display even
unsupported settings in their output if needed.
lvmconfig --type list displays plain list of configuration settings.
Some of the existing decorations can be used (--withsummary and
--withversions) as well as existing options/switches (--ignoreadvanced,
--ignoreunsupported, --ignorelocal, --atversion).
For example (displaying only "config" section so the list is not long):
$lvmconfig --type list config
config/checks
config/abort_on_errors
config/profile_dir
$ lvmconfig --type list --withsummary config
config/checks - If enabled, any LVM configuration mismatch is reported.
config/abort_on_errors - Abort the LVM process if a configuration mismatch is found.
config/profile_dir - Directory where LVM looks for configuration profiles.
$ lvmconfig -l config
config/checks - If enabled, any LVM configuration mismatch is reported.
config/abort_on_errors - Abort the LVM process if a configuration mismatch is found.
config/profile_dir - Directory where LVM looks for configuration profiles.
$ lvmconfig --type list --withsummary --withversions config
config/checks - If enabled, any LVM configuration mismatch is reported. [2.2.99]
config/abort_on_errors - Abort the LVM process if a configuration mismatch is found. [2.2.99]
config/profile_dir - Directory where LVM looks for configuration profiles. [2.2.99]
Example with --atversion (displaying global section):
$ lvmconfig --type list global
global/umask
global/test
global/units
global/si_unit_consistency
global/suffix
global/activation
global/fallback_to_lvm1
global/format
global/format_libraries
global/segment_libraries
global/proc
global/etc
global/locking_type
global/wait_for_locks
global/fallback_to_clustered_locking
global/fallback_to_local_locking
global/locking_dir
global/prioritise_write_locks
global/library_dir
global/locking_library
global/abort_on_internal_errors
global/detect_internal_vg_cache_corruption
global/metadata_read_only
global/mirror_segtype_default
global/raid10_segtype_default
global/sparse_segtype_default
global/lvdisplay_shows_full_device_path
global/use_lvmetad
global/thin_check_executable
global/thin_dump_executable
global/thin_repair_executable
global/thin_check_options
global/thin_repair_options
global/thin_disabled_features
global/cache_check_executable
global/cache_dump_executable
global/cache_repair_executable
global/cache_check_options
global/cache_repair_options
global/system_id_source
global/system_id_file
$ lvmconfig --type list global --atversion 2.2.50
global/umask
global/test
global/units
global/suffix
global/activation
global/fallback_to_lvm1
global/format
global/format_libraries
global/segment_libraries
global/proc
global/locking_type
global/wait_for_locks
global/fallback_to_clustered_locking
global/fallback_to_local_locking
global/locking_dir
global/library_dir
global/locking_library
Example:
/dev/loop0 and /dev/loop1 are duplicates,
created by copying one backing file to the
other.
'identity /dev/loopX' creates an identity
mapping for loopX named idmloopX, which
adds a duplicate for the named device.
The duplicate selection code for lvmetad is
incomplete, and lvmetad is disabled for this
example.
[~]# losetup -f loopfile0
[~]# pvs
PV VG Fmt Attr PSize PFree
/dev/loop0 foo lvm2 a-- 308.00m 296.00m
[~]# losetup -f loopfile1
[~]# pvs
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/loop1 not /dev/loop0
Using duplicate PV /dev/loop1 which is more recent, replacing /dev/loop0
PV VG Fmt Attr PSize PFree
/dev/loop1 foo lvm2 a-- 308.00m 308.00m
[~]# ./identity /dev/loop0
[~]# pvs
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/loop1 not /dev/loop0
Using duplicate PV /dev/loop1 without holders, replacing /dev/loop0
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/mapper/idmloop0 not /dev/loop1
Using duplicate PV /dev/mapper/idmloop0 from subsystem DM, replacing /dev/loop1
PV VG Fmt Attr PSize PFree
/dev/mapper/idmloop0 foo lvm2 a-- 308.00m 296.00m
[~]# ./identity /dev/loop1
[~]# pvs
WARNING: duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV is being used from both devices /dev/loop0 and /dev/loop1
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/loop1 not /dev/loop0
Using duplicate PV /dev/loop1 which is more recent, replacing /dev/loop0
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/mapper/idmloop0 not /dev/loop1
Using duplicate PV /dev/mapper/idmloop0 from subsystem DM, replacing /dev/loop1
Found duplicate PV LnSOEqzEYED3RvIOa5PZP2s7uyuBLmAV: using /dev/mapper/idmloop1 not /dev/mapper/idmloop0
Using duplicate PV /dev/mapper/idmloop1 which is more recent, replacing /dev/mapper/idmloop0
PV VG Fmt Attr PSize PFree
/dev/mapper/idmloop1 foo lvm2 a-- 308.00m 308.00m
Describe
thin_check_options, thin_repair_options,
cache_check_option, scache_repair_options
as a "list of options", rather than a "string of options"
because a single string, e.g. "-q --clear-needs-check-flag"
does not work, and needs to be entered as a list,
e.g. ["-q", "--clear-needs-check-flag"]
The settings which have their default value evaluated in runtime should
have their 'unconfigured' counterparts also evaluated in runtime since
those values can be constructed by using other settings.
For example, before this patch:
$ lvm dumpconfig --type default --unconfigured devices/cache_dir devices/cache
cache_dir="@DEFAULT_SYS_DIR@/@DEFAULT_CACHE_SUBDIR@"
cache="/etc/lvm/cache/.cache
With this patch applied:
$ lvm dumpconfig --type default --unconfigured devices/cache_dir devices/cache
cache_dir="@DEFAULT_SYS_DIR@/@DEFAULT_CACHE_SUBDIR@"
cache="@DEFAULT_SYS_DIR@/@DEFAULT_CACHE_SUBDIR@/.cache"
The @something@ used for unconfigured default value is not bound to
CFG_TYPE_STRING settings defined in config_settings.h, it can be
used for any other config type too.
Show full chain of ancestors and descendants for snapshots
(both thick and thin - in case of thick, the "ancestor" field
is actually equal to "origin" field as snapshots can't be
chained for thick snapshots).
These fields display current state as it is, they do not
display any history! If the snapshot chain is broken in
the middle, we don't report the historical origin (this
is going to be a part of another patch and a different
set of fields or just a switch for existing fields to
show ancestors and descendants with history included).
For example:
(origin --> snapshot)
lvol1 --> lvol2 --> lvol3 --> lvol4
\
--> lvol5 --> lvol6 --> lvol7 --> lvol8
$ lvs -o name,pool_lv,origin,ancestors,descendants vg
LV Pool Origin Ancestors Descendants
lvol1 pool lvol2,lvol3,lvol4,lvol5,lvol6,lvol7,lvol8
lvol2 pool lvol1 lvol1 lvol3,lvol4,lvol5,lvol6,lvol7,lvol8
lvol3 pool lvol2 lvol2,lvol1 lvol4
lvol4 pool lvol3 lvol3,lvol2,lvol1
lvol5 pool lvol2 lvol2,lvol1 lvol6,lvol7,lvol8
lvol6 pool lvol5 lvol5,lvol2,lvol1 lvol7,lvol8
lvol7 pool lvol6 lvol6,lvol5,lvol2,lvol1 lvol8
lvol8 pool lvol7 lvol7,lvol6,lvol5,lvol2,lvol1
Scenario:
$ vgs -o+vg_mda_copies
VG #PV #LV #SN Attr VSize VFree #VMdaCps
fedora 1 2 0 wz--n- 9.51g 0 unmanaged
vg 16 9 0 wz--n- 1.94g 1.83g 2
$ lvs -o+read_ahead vg/lvol6 vg/lvol7
LV VG Attr LSize Pool Origin Data% Rahead
lvol6 vg Vwi-a-tz-- 1.00g pool lvol5 0.00 auto
lvol7 vg Vwi---tz-k 1.00g pool lvol6 256.00k
Before this patch:
$vgs -o vg_name,vg_mda_copies -S 'vg_mda_copies < unmanaged'
VG #VMdaCps
vg 2
Problem:
Reserved values can be only used with exact match = or !=, not <,<=,>,>=.
In the example above, the "unamanaged" is internally represented as
18446744073709551615, but this should be ignored while not comparing
field directly with "unmanaged" reserved name with = or !=. Users
should not be aware of this internal mapping of the reserved value
name to its internal value and hence it doesn't make sense for such
reserved value to take place in results of <,<=,> and >=.
There's no order defined for reserved values!!! It's a special
*reserved* value that is taken out of the usual value range
of that type.
This is very similar to what we have already fixed with
2f7f6932dc, but it's the other way round
now - we're using reserved value name in selection criteria now
(in the patch 2f7f693, we had concrete value and we compared it
with the reserved value). So this patch completes patch 2f7f693.
This patch also fixes this problem:
$ lvs -o+read_ahead vg/lvol6 vg/lvol7 -S 'read_ahead > 32k'
LV VG Attr LSize Pool Origin Data% Rahead
lvol6 vg Vwi-a-tz-- 1.00g pool lvol5 0.00 auto
lvol7 vg Vwi---tz-k 1.00g pool lvol6 256.00k
Problem:
In the example above, the internal reserved value "auto" is in the
range of selection "> 32k" - it shouldn't match as well. Here the
"auto" is internally represented as MAX_DBL and of course, numerically,
MAX_DBL > 256k. But for users, the reserved value should be uncomparable
to any number so the mapping of the reserved value name to its interna
value is transparent to users. Again, there's no order defined for
reserved values and hence it should never match if using <,<=,>,>=
operators.
This is actually exactly the same problem as already described in
2f7f6932dc, but that patch failed for
size field types because of incorrect internal representation used.
With this patch applied, both problematic scenarios mentioned
above are fixed now:
$ vgs -o vg_name,vg_mda_copies -S 'vg_mda_copies < unmanaged'
(blank)
$ lvs -o+read_ahead vg/lvol6 vg/lvol7 -S 'read_ahead > 32k'
LV VG Attr LSize Pool Origin Rahead
lvol7 vg Vwi---tz-k 1.00g pool lvol6 256.00k
By default these are empty strings, so the config settings
should be flagged as undefined, so they will be commented
out of the generated config. Otherwise, the lines:
thin_repair_options=""
cache_repair_options=""
in the dump output cause a warning when processed since
lvm doesn't want an empty string.
Also regenerate lvm.conf.in.
Rename envvar LVM_LOG_FILE_UNLINK_STATUS to LVM_EXPECTED_EXIT_STATUS
and change compare sign from '!' to '>'.
Validate LVM_LOG_FILE_EPOCH and support strictly only
up-to 32 alpha chars. If the content doesn't pass
epoch is simply ignored.
Add support for 2 new envvars for internal lvm2 test suite
(though it could be possible usable for other cases)
LVM_LOG_FILE_EPOCH
Whether to add 'epoch' extension that consist from
the envvar 'string' + pid + starttime in kernel units
obtained from /proc/self/stat.
LVM_LOG_FILE_UNLINK_STATUS
Whether to unlink the log depending on return status value,
so if the command is successful the log is automatically
deleted.
API is still for now experimental to catch various issue.
--withfullcomments prints all comment lines for each config option.
--withcomments prints only the first comment line, which should be
a short one-line summary of the option.
When performing initial allocation (so there is nothing yet to
cling to), use the list of tags in allocation/cling_tag_list to
partition the PVs. We implement this by maintaining a list of
tags that have been "used up" as we proceed and ignoring further
devices that have a tag on the list.
https://bugzilla.redhat.com/983600
Add A_PARTITION_BY_TAGS set when allocated areas should not share tags
with each other and allow _match_pv_tags to accept an alternative list
of tags. (Not used yet.)
Comments from the sample config files are copied into
the comment field of the config settings structure.
This includes only minimal changes to the text.
With this in place, the sample config files can
be generated from 'lvm dumpconfig', and content
for an lvm.conf man page can also be generated.
pv_write is called both to write orphans and to rewrite PV headers
of PVs in VGs. It needs to select the correct VG id so that the
internal cache state gets updated correctly.
It only affected commands that involved further steps after
the pv_write and was often masked because the metadata would
be re-read off disk and correct itself.
"Incorrect metadata area header checksum" warnings appeared.
Example:
Create vg1 containing dev1, dev2 and dev3.
Hide dev1 and dev2 from the system.
Fix up vg1 with vgreduce --removemissing.
Bring back dev1 and dev2.
In a single operation reinstate dev1 and dev2 into vg1 (vgextend).
Done as separate operations (automatically fix-up dev1 and dev2 as orphans,
then vgextend) it worked, but done all in one go the internal cache got
corrupted and warnings about checksum errors appeared.
Commit 80f4b4b803
introduced undesirable side-effects for lvm2app user
which happens to be our own python binding.
It appear obtaing pvs list keeps global lock.
So restricting this to VG_GLOBAL READ locks and skip
the drop skip if WRITE lock is held.
Do not keep dangling LVs if they're removed from the vg->lvs list and
move them to vg->removed_lvs instead (this is actually similar to already
existing vg->removed_pvs list, just it's for LVs now).
Once we have this vg->removed_lvs list indexed so it's possible to
do lookups for LVs quickly, we can remove the LV_REMOVED flag as
that one won't be needed anymore - instead of checking the flag,
we can directly check the vg->removed_lvs list if the LV is present
there or not and to say if the LV is removed or not then. For now,
we don't have this index, but it may be implemented in the future.
This avoids a problem in which we're using selection on LV list - we
need to do the selection on initial state and not on any intermediary
state as we process LVs one by one - some of the relations among LVs
can be gone during this processing.
For example, processing one LV can cause the other LVs to lose the
relation to this LV and hence they're not selectable anymore with
the original selection criteria as it would be if we did selection
on inital state. A perfect example is with thin snapshots:
$ lvs -o lv_name,origin,layout,role vg
LV Origin Layout Role
lvol1 thin,sparse public,origin,thinorigin,multithinorigin
lvol2 lvol1 thin,sparse public,snapshot,thinsnapshot
lvol3 lvol1 thin,sparse public,snapshot,thinsnapshot
pool thin,pool private
$ lvremove -ff -S 'lv_name=lvol1 || origin=lvol1'
Logical volume "lvol1" successfully removed
The lvremove command above was supposed to remove lvol1 as well as
all its snapshots which have origin=lvol1. It failed to do so, because
once we removed the origin lvol1, the lvol2 and lvol3 which were
snapshots before are not snapshots anymore - the relations change
as we're processing these LVs one by one.
If we do the selection first and then execute any concrete actions on
these LVs (which is what this patch does), the behaviour is correct
then - the selection is done on the *initial state*:
$ lvremove -ff -S 'lv_name=lvol1 || origin=lvol1'
Logical volume "lvol1" successfully removed
Logical volume "lvol2" successfully removed
Logical volume "lvol3" successfully removed
Similarly for all the other situations in which relations among
LVs are being changed by processing the LVs one by one.
This patch also introduces LV_REMOVED internal LV status flag
to mark removed LVs so they're not processed further when we
iterate over collected list of LVs to be processed.
Previously, when we iterated directly over vg->lvs list to
process the LVs, we relied on the fact that once the LV is removed,
it is also removed from the vg->lvs list we're iterating over.
But that was incorrect as we shouldn't remove LVs from the list
during one iteration while we're iterating over that exact list
(dm_list_iterate_items safe can handle only one removal at
one iteration anyway, so it can't be used here).
The code never mixes reads of committed and precommitted metadata,
so there's no need to attempt to set PRECOMMITTED when
*use_previous_vg is being set.
Refactor the recent metadata-reading optimisation patches.
Remove the recently-added cache fields from struct labeller
and struct format_instance.
Instead, introduce struct lvmcache_vgsummary to wrap the VG information
that lvmcache holds and add the metadata size and checksum to it.
Allow this VG summary information to be looked up by metadata size +
checksum. Adjust the debug log messages to make it clear when this
shortcut has been successful.
(This changes the optimisation slightly, and might be extendable
further.)
Add struct cached_vg_fmtdata to format-specific vg_read calls to
preserve state alongside the VG across separate calls and indicate
if the details supplied match, avoiding the need to read and
process the VG metadata again.
Fixes segfault when 'pvs' encounters two different PVs sharing the same
uuid but one an orphan, the other in a VG.
If VG_GLOBAL is held, there seems no point in doing a full scan more
than once.
If undesirable side-effects show up, we can try restricting this to
VG_GLOBAL READ locks. The original code dates back to 2.02.40.
When pvscan --cache --major --minor command is issued from
udev REMOVE event, it basically resulted into a whole device
scan since the device was missing. So avoid such scan
and first check via /sysfs (when available) if such device actually
exists.
When available use nanosecond stat info.
If commands are running closely enough after config update,
the .cache file from persistent filter could have been ignored.
This happens sometimes during i.e. synthetic test suite run.
Metadata areas which are marked as ignored should not be scanned
and read during pvscan --cache. Otherwise, this can cause lvmetad
to cache out-of-date metadata in case other PVs with fresh metadata
are missing by chance.
Make this to work like in non-lvmetad case where the behaviour would
be the same as if the PV was orphan (in case we have no other PVs
with valid non-ignored metadata areas).
When lvm1 PVs are visible, and lvmetad is used, and the foreign
option was included in the reporting command, the reporting
command would fail after the 'pvscan all devs' function saw
the lvm1 PVs. There is no reason the command should fail
because of the lvm1 PVs; they should just be ignored.
Return 1 on success in pvdisplay_short() and lvdisplay_full()
so commands like vgdisplay are not printinig stracktraces
on successful passes.
As the results of fail/success have been internally ignored for those
calls, it had no other visible side effect - command's return value was
still 0 (success).
Detect an lvm1 system id by looking at the WRITE_LOCKED flag.
Don't copy this lvm1 system id into vg->system_id so that the
restrictions associated with the new system id are not applied
to the old VG with the inherited lvm1 system id.
Since we take a lock inside vg_lock_newname() and we do a full
detection of presence of vgname inside all scanned labels,
there is no point to do this for second time to be sure
there is no such vg.
The only side-effect of such call would be a full validation of
some already exising VG metadata - but that's not the task for
vgcreate when create a new VG.
This call noticable reduces number of scans during 'vgcreate'.
Use similar logic as with text_vg_import_fd() and avoid repeated
parsing of same mda and its config tree for vgname_from_mda().
Remember last parsed vgname, vgid and creation_host in labeller
structure and if the metadata have the same size and checksum,
return this stored info.
TODO: The reuse of labeller struct is not ideal, some lvmcache API for
this functionality would be nicer.
When reading VG mda from multiple PVs - do all the validation only
when mda is seen for the first time and when mda checksum and length
is same just return already existing VG pointer.
(i.e. using 300PVs for a VG would lead to create and destroy 300 config trees....)
Previous versions of lvm will not obey the restrictions
imposed by the new system_id, and would allow such a VG
to be written. So, a VG with a new system_id is further
changed to force previous lvm versions to treat it as
read-only. This is done by removing the WRITE flag from
the metadata status line of these VGs, and putting a new
WRITE_LOCKED flag in the flags line of the metadata.
Versions of lvm that recognize WRITE_LOCKED, also obey the
new system_id. For these lvm versions, WRITE_LOCKED is
identical to WRITE, and the rules associated with matching
system_id's are imposed.
A new VG lock_type field is also added that causes the same
WRITE/WRITE_LOCKED transformation when set. A previous
version of lvm will also see a VG with lock_type as read-only.
Versions of lvm that recognize WRITE_LOCKED, must also obey
the lock_type setting. Until the lock_type feature is added,
lvm will fail to read any VG with lock_type set and report an
error about an unsupported lock_type. Once the lock_type
feature is added, lvm will allow VGs with lock_type to be
used according to the rules imposed by the lock_type.
When both system_id and lock_type settings are removed, a VG
is written with the old WRITE status flag, and without the
new WRITE_LOCKED flag. This allows old versions of lvm to
use the VG as before.
The seg_monitor did not display monitored status for thick snapshots
and mirrors (with mirror log *not* mirrored). The seg monitor did work
correctly even before for other segtypes - thins and raids.
Before (mirrors and snapshots, only mirrors with mirrored log properly displayed monitoring status):
[0] f21/~ # lvs -a -o lv_name,lv_layout,lv_role,seg_monitor vg
LV Layout Role Monitor
mirror mirror public
[mirror_mimage_0] linear private,mirror,image
[mirror_mimage_1] linear private,mirror,image
[mirror_mlog] linear private,mirror,log
mirror_with_mirror_log mirror public monitored
[mirror_with_mirror_log_mimage_0] linear private,mirror,image
[mirror_with_mirror_log_mimage_1] linear private,mirror,image
[mirror_with_mirror_log_mlog] mirror private,mirror,log monitored
[mirror_with_mirror_log_mlog_mimage_0] linear private,mirror,image
[mirror_with_mirror_log_mlog_mimage_1] linear private,mirror,image
thick_origin linear public,origin,thickorigin
thick_snapshot linear public,snapshot,thicksnapshot
With this patch applied (monitoring status displayed for all mirrors and snapshots):
[0] f21/~ # lvs -a -o lv_name,lv_layout,lv_role,seg_monitor vg
LV Layout Role Monitor
mirror mirror public monitored
[mirror_mimage_0] linear private,mirror,image
[mirror_mimage_1] linear private,mirror,image
[mirror_mlog] linear private,mirror,log
mirror_with_mirror_log mirror public monitored
[mirror_with_mirror_log_mimage_0] linear private,mirror,image
[mirror_with_mirror_log_mimage_1] linear private,mirror,image
[mirror_with_mirror_log_mlog] mirror private,mirror,log monitored
[mirror_with_mirror_log_mlog_mimage_0] linear private,mirror,image
[mirror_with_mirror_log_mlog_mimage_1] linear private,mirror,image
thick_origin linear public,origin,thickorigin
thick_snapshot linear public,snapshot,thicksnapshot monitored
If configuration setting is marked in config_setting.h with CFG_DISABLED
flag, default value is always used for such setting, no matter if it's defined
by user (in --config/lvm.conf/lvmlocal.conf).
A warning message is displayed if this happens:
For example:
[1] f21/~ # lvm dumpconfig --validate
WARNING: Configuration setting global/system_id_source is disabled. Using default value.
LVM configuration valid.
[1] f21/~ # pvs
WARNING: Configuration setting global/system_id_source is disabled. Using default value.
PV VG Fmt Attr PSize PFree
/dev/sdb lvm2 --- 128.00m 128.00m
...
Set ACCESS_NEEDS_SYSTEM_ID VG status flag whenever there is
a non-lvm1 system_id set. Prevents concurrent access from
older LVM2 versions.
Not set on VGs that bear a system_id only due to conversion
from lvm1 metadata.
Export _lvm1_system_id as generate_lvm1_system_id and call it in
vg_setup() so it is set before writing the metadata to disk
and not missing from the initial metadata backup file.
format_text processes both lvm2 on-disk metadata and metadata read
from other sources such as backup files. Add original_fmt field
to retain the format type of the original metadata.
Before this patch, /etc/lvm/archives would contain backups of
lvm1 metadata with format = "lvm2" unless the source was lvm1 on-disk
metadata.
The vg->lvm1_systemd_id needs to be initialized as all the code around
counts with that. Just like we initialize lvm1_system_id in vg_create
(no matter if it's actually LVM1 or LVM2 format), this patch adds this
init in alloc_vg as well so the rest of the code does not segfaul
when trying to access vg->lvm1_system_id.
In log messages refer to it as system ID (not System ID).
Do not put quotes around the system_id string when printing.
On the command line use systemid.
In code, metadata, and config files use system_id.
In lvmsystemid refer to the concept/entity as system_id.
The only realistic way for a host to have active LVs in a
foreign VG is if the host's system_id (or system_id_source)
is changed while LVs are active.
In this case, the active LVs produce an warning, and access
to the VG is implicitly allowed (without requiring --foreign.)
This allows the active LVs to be deactivated.
In this case, rescanning PVs for the VG offers no benefit.
It is not possible that rescanning would reveal an LV that
is active but wasn't previously in the VG metadata.
cmirror uses the CPG library to pass messages around the cluster and maintain
its bitmaps. When a cluster mirror starts-up, it must send the current state
to any joining members - a checkpoint. When mirrors are large (or the region
size is small), the bitmap size can exceed the message limit of the CPG
library. When this happens, the CPG library returns CPG_ERR_TRY_AGAIN.
(This is also a bug in CPG, since the message will never be successfully sent.)
There is an outstanding bug (bug 682771) that is meant to lift this message
length restriction in CPG, but for now we work around the issue by increasing
the mirror region size. This limits the size of the bitmap and avoids any
issues we would otherwise have around checkpointing.
Since this issue only affects cluster mirrors, the region size adjustments
are only made on cluster mirrors. This patch handles cluster mirror issues
involving pvmove, lvconvert (from linear to mirror), and lvcreate. It also
ensures that when users convert a VG from single-machine to clustered, any
mirrors with too many regions (i.e. a bitmap that would be too large to
properly checkpoint) are trapped.
A foreign VG should be silently ignored by a reporting/display
command like 'vgs'. If the reporting/display command specifies
a foreign VG by name on the command line, it should produce an
error message.
Scanning commands pvscan/vgscan/lvscan are always allowed to
read and update caches from all PVs, including those that belong
to foreign VGs.
Other non-report/display/scan commands always ignore a foreign
VG, or report an error if they attempt to use a foreign VG.
vgimport should always invalidate the lvmetad cache because
lvmetad likely holds a pre-vgexported copy of the VG.
(This is unrelated to using foreign VGs; the pre-vgexported
VG may have had no system_id at all.)
When checking whether the system ID permits access to a VG, check for
each permitted situation first, and only then issue the appropriate
error message. Always issue a message for now. (We'll try to
suppress some of those later when the VG concerned wasn't explicitly
requested.)
Add more messages to try to ensure every return code is checked and
every error path (and only an error path) contains a log_error().
Add self-correction to vgchange -c to deal with situations where
the cluster state and system ID state are out-of-sync (e.g. if
old tools were used).
Move the lvm1 sys ID into vg->lvm1_system_id and reenable the #if 0
LVM1 code. Still display the new-style system ID in the same
reporting field, though, as only one can be set.
Add a format feature flag FMT_SYSTEM_ON_PVS for LVM1 and disallow
access to LVM1 VGs if a new-style system ID has been set.
Treat the new vg->system_id as const.
Allow cmd->unknown_system_id to be cleared during toolcontext
refresh.
Set a default value of "none" for global/system_id_source.
Allow local/system_id to be empty so it's not impossible for
a later config file to remove it.
In a file containing a system ID:
Any whitespace at the start of a line is ignored;
Blank lines are ignored;
Any characters after a # are ignored along with the #.
The system ID is obtained by processing the first line with
non-ignored characters.
If further lines with non-ignored characters follow, a warning is
issued.
Add WARNING messages if there are problems setting the requested
system ID.
Ban "localhost" as a prefix regardless of the system_id_source.
Use cmd->hostname instead of calling uname again.
Make system_id_source values case-insensitive (as with new settings like
log_debug_classes) and also accept machine-id to match the filename.
Require system ID to begin with an alphanumeric character.
Rename fn to make clear it's only validation for systemid
and always terminate result rather than imposing this on the caller.
In 2.02.99, _init_tags() inadvertently began to ignore the
dm_config_tree struct passed to it. "tags" sections are not
merged together, so the "tags" section in the main config file was
being processed repeatedly and other "tags" sections were ignored.
Dop unused value assignments.
Unknown is detected via other combination
(!linear && !striped).
Also change the log_error() message into a warning,
since the function is not really returning error,
but still keep the INTERNAL_ERROR.
Ret value is always set later.
The dev ext source must be reset for the dev_cache_get call
(which evaluates filters), not lvmcache_label_scan - so fix
original commit 727c7ff85d.
Also, add comments in _pvcreate_check fn explaining why
refresh filter and rescan is needed and exactly in which
situations.
We exclude some signatures from being wiped when using blkid wiping.
These are signatures which we simply overwrite. For example, the
LVM2_member signature which denotes a PV - if we call pvcreate on
existing PV, we just overwrite the PV header, no need to wipe it.
Previously, we counted such signatures as if they were wiped
and they were counted in the final number of wiped signatures
that _wipe_known_signatures_with_blkid fn returned in the "wiped"
output arg. Then the code checking this output arg could be
mislead that wiping happened while no wiping took place in real
and we could fire some code uselessly based on this information
(e.g. refreshing filters/rescanning - see also
commit 6b4066585f).
Before, we refreshed filters and we did full rescan of devices if
we passed through wiping (wipe_known_signatures fn call). However,
this fn returns success even if no signatures were found and so
nothing was wiped. In this case, it's not necessary to do the
filter refresh/rescan of devices as nothing changed clearly.
This patch exports number of wiped signatures from all the
wiping functions below. The caller (_pvcreate_check) then checks
whether any wiping was done at all and if not, no refresh/rescan
is done, saving some time and resources.
pvcreate code path executes signature wiping if there are any signatures
found on device to prepare the device for PV. When the signature is wiped,
the WATCH udev rule triggers the event which then updates udev database
with fresh info, clearing the old record about previous signature.
However, when we're using udev db as dev-ext source, we'd need to wait
for this WATCH-triggered event. But we can't synchronize against such
events (at least not at this moment). Without this sync, if the code
continues, the device could still be marked as containing the old
signature if reading udev db. This may end up even with the device
to be still filtered, though the signature is already wiped.
This problem is then exposed as (an example with md components):
$ mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sda /dev/sdb --run
$ mdadm -S /dev/md0
$ pvcreate -y /dev/sda
Wiping linux_raid_member signature on /dev/sda.
/dev/sda: Couldn't find device. Check your filters?
$ echo $?
5
So we need to temporarily switch off "udev" dev-ext source here
in this part of pvcreate code until we find a way how to sync
with WATCH events.
(This problem does not occur with signature wiping which we do
on newly created LVs since we already handle this properly with
our udev flags - the LV_NOSCAN/LV_TEMPORARY flag. But we can't use
this technique for non-dm devices to keep WATCH rule under control.)
It's cleaner this way - do not mix static and dynamic
(init_processing_handle) initializers. Use the dynamic one everywhere.
This makes it easier to manage the code - there are no "exceptions"
then and we don't need to take care about two ways of initializing the
same thing - just use one common initializer throughout and it's clear.
Also, add more comments, mainly in the report_for_selection fn explaining
what is being done and why with respect to the processing_handle and
selection_handle.
When lvm2 is build with valgrind pool detection - always disable
memcheck, since pool memory allocation are unconditionaly passed
into valgrind library.
This makes a difference when using selection criteria based on
these fields - if those fields are defined as DM_REPORT_FIELD_TYPE_SIZE
(in contrast to DM_REPORT_FIELD_TYPE_NUMBER), units are also
recognize in selection clause.
For example:
$ lvs -o+seg_start vg1/lv2
LV VG Attr LSize Start
lv2 vg1 -wi-a----- 12.00m 0
lv2 vg1 -wi-a----- 12.00m 8.00m
Before this patch:
$ lvs -o+seg_start --select 'seg_start=8m'
Found size unit specifier but numeric value expected for selection field seg_start.
Selection syntax error at 'seg_start=8m'.
Use 'help' for selection to get more help.
With this patch applied:
$lvs -o+seg_start --select 'seg_start=8m'
LV VG Attr LSize Start
lv2 vg1 -wi-a----- 12.00m 8.00m
(the same applies for ba_start and vg_free fields)
The report_for_selection does the actual "reporting for selection only".
The selection status will be saved in struct selection_handle's "selected"
variable.
The new "report_init_for_selection" is just a wrapper over
dm_report_init_with_selection that initializes reporting for selection
only. This means we're not going to do the actual reporting to output
for display and as such we intialize reporting as if no fields are reported
or sorted. The only fields "reported" are taken from the selection criteria
string and all such fields are marked as hidden automatically (FLD_HIDDEN flag).
These fields are used solely for selection criteria matching.
Also, modify existing report_object function that was used for reporting to
output for display. Now, it can either cause reporting to output or reporting
for selection only. The selection result is stored in struct selection_handle's
"selected" variable which can be handled further by any report_object caller.
This patch replaces "void *handle" with "struct processing_handle *handle"
in process_each_*, process_single_* and related functions.
The struct processing_handle consists of two handles inside now:
- the "struct selection_handle *selection_handle" used for
applying selection criteria while processing process_each_*,
process_single_* and related functions (patches using this
logic will follow)
- the "void* custom_handle" (this is actually the original handle
used before this patch - a pointer to custom data passed into
process_each_*, process_single_* and related functions).
Older lvm2 tools where always providing linear mapping for thin pool.
Recent lvm2 version however support external usage of thin pool and
empty/unused pools are loaded without such external linear mapping.
So this patch covers 'upgrade' problem, where older tool has activated
thin-pool with 'linear' layer mapping, and newer tools didn't expected
such mapping to exist and were not able to deactivate such table.
So before checking for new layout in dm-table, check if there is not
an old one already there.
Just like MD filtering that detects components of software RAID (md),
add detection for firmware RAID.
We're not adding any native code to detect this - there are lots of
firmware RAIDs out there which is just out of LVM scope. However,
with current changes with which we're able to get device info from
external sources (e.g. external_device_info_source="udev"), we can
do this easily if the external device status source has this kind
of information - which is the case of "udev" source where the results
of blkid scans are stored.
This detection should cover all firmware RAIDs that blkid can detect and
which are identified as:
ID_FS_TYPE = {adaptec,ddf,hpt45x,hpt37x,isw,jmicron,lsi_mega,nvidia,promise_fasttrack,silicon_medley,via}_raid_member
Partitioned devices are marked in udev db as:
ID_PART_TABLE="<partition table type name>"
and at the same time they are *not* marked with:
ID_PART_ENTRY_DISK="<parent disk major:minor>"
Where partition table type name is dos/gpt/... But checking the presence
of this variable is enough for LVM here - it just needs to know whether
there's a partition table or not, not interested in the actual type.
The same applies for parent disk major:minor.
The filter-partitioned code should contain only checks in "partition" domain.
The check for pv_min_size should actually be a part of filter-usable.
If the device size is less than pv_min_size, such device is not usable
as a PV so this check clearly belongs here logically.
With udev external info source, we can get device size via libudev's
sysfs reading interface and we can avoid opening the device this way
effectively.
mpath components are marked in udev db as:
ID_FS_TYPE="mpath_member"
or
DM_MULTIPATH_DEVICE_PATH="1"
(it depends on udev rule/blkid version used for handling mpath)
Composite filter is a filter that can put several filters in one set.
This patch adds a switch when creating the composite filter which will
enable or disable external device info handles for all the filters
the composite filter encompasses.
We want to use this external device info for majority of the filters
which are in the "lvmetad filter chain" (or the respective part if
we're not using lvmetad).
Following patches will use the enabled external device handle in
concrete filters from the composite filter...
for_each_sub_lv() now scans in depth also pools, however for
rename we actually do want to skip pools.
So add a new for_each_sub_lv_except_pools() to be used by rename,
every other user of for_each_sub_lv() scans every sub LV with pools
included.
This is i.e. necessary for properly working preload of pools
that are using raid arrays.
LVSINFO, LVSSTATUS and LVSINFOSTATUS is the same as LVS, just with some
extra info/status decoration attached to it. Recognize this when looking
for properties for lvm2app. This fixes lvm_lv_get_property lvm2app call
for fields which already use LVS{INFO,STATUS,INFOSTATUS} - currently,
this is lv_attr field which was converted to LVSINFOSTATUS from
pure LVS type.
This is a regression from v115 where some of the fields/properties
were converted to using the common "struct lvinfo" and
"struct lv_seg_status" so we don't need to issue info and status
ioctl several times per one reported line. Not all fields are
converted yet, but one that *is* converted is the lv_attr field
with the lv_attr_dup counterpart used in lvm_lv_get_attr lvm2app fn.
These changes were introduced with e34b004422
and later - this patch introduced the "info_ok" field in the
lv_with_info_and_seg_status structure which encapsulates the lvinfo
and lv_seg_status struct.
For the lv_attr_dup, the lv_attr_dup code missed the
assignment for the "info_ok" flag which saves the result of the
lv_info_with_seg_status call. Hence such info was marked
as unusable - unknown and it was returned as such via lvm_lv_get_attr
lvm2app fn.
When cache_mode is undefined, the read of metadata will miss to
set a bit with mode and fails to process metadata on internal
error:
Internal error: LV vg/lvol1 has uknown feature flags 0.
Fix it by setting it to writethrough mode.
Check splitted leg is active before preload.
(Since splitmirrors currently only does work active raid volumes
it's not a change for current code flow).
Minor optimization included - when already positively checked
for raid image don't check again for raid metadata.
for_each_sub_lv() normally does not put pool_lv into deps.
So for now go around it in 'lv_preload()' and add explicit
call with pool.
TODO: think about a better way, we want pool_lv deps only in certain
moments, so maybe for_each_sub_lv() needs new arg for this.
When raid leg is extracted, now the preload code handles this state
correctly and put proper new table entry into dm tree,
so the activation of extracted leg and removed metadata works
after commit.
When raid is being splitted, extracted leg & metadata
is still floating in the table - and thus we need to
detect this case and properly preload their matching
table so consequent activation of extracted LVs properly
renames (and FREES) existing raid images, so ongoing
image name shifting will work.
For example, with dmeventd/executable set to "" which is not allowed for
this setting, the config validation now ends up with:
$ lvm dumpconfig --validate
Configuration setting "dmeventd/executable" invalid. It cannot be set to an empty value.
LVM configuration invalid.
This check for empty values for string config settings was not
done before (we only checked empty arrays, but not scalar strings).
Rename original lv_error_when_full field to lv_when_full and also
convert it from binary field to string field displaying three
possible values: "error", "queueu" or "" (blank for undefined).
$ lvs vg/pool vg/pool1 vg/linear_lv -o+lv_when_full
LV VG Attr LSize Data% Meta% WhenFull
linear_lv vg -wi-a----- 4.00m
pool vg twi-aotz-- 4.00m 0.00 0.98 queue
pool1 vg twi-a-tz-- 4.00m 0.00 0.88 error
For -S|--select these synonyms are recognized:
"error" -> "error when full", "error if no space"
"queue" -> "queue when full", "queue if no space"
"" -> "undefined"
Add separate LVSINFOSTATUS field type for fields which display both
dm info-like and dm status-like information.
The internal interface is there with the introduction of LVSSTATUS
field type which can cope with the combination of LVSSTATUS
and LVSINFO field types (several fields).
However, till now, we considered that *single* field can display
either LVSINFO or LVSSTATUS, but not both at the same time.
Till now, we haven't had single field which needs both - hence
add LVSINFOSTATUS field type for such fields as we currently
need this for the lv_attr field which requires combination of
info and status.
This patch just adds interface for an ability to register such fields
(the code that copes with this is already in).
Recently the single 'status' code has been used for number of cache
features.
Extend the API a little bit to allow usage also for lv_attr_dup.
As the function itself is used in lvm2api - add a new function:
lv_attr_dup_with_info_and_seg_status() that is able to use
grabbed info & status information.
report_init() is now using directly passed lvdm struct pointer
which holds the infomation whether lv_info() was correctly obtained or
there was some error when trying to read it.
Move 'healt' attribute to status.
TODO convert raid function to use the already known status.
The previous patch felt short WRT disabling allocation on PVs holding other
legs of the RAID LV persistently; this patch introduces an internal,
transient PV flag PV_ALLOCATION_PROHIBITED to address this very problem.
General problem description for completeness:
An 'lvconvert --repair $RAID_LV" to replace a failed leg of a multi-segment
RAID10/4/5/6 logical volume can lead to allocation of (parts of) the replacement
image component pair on the physical volume of another image component
(e.g. image 0 allocated on the same PV as image 1 silently impeding resilience).
Patch fixes this severe resilince issue by prohibiting allocation on PVs
already holding other legs of the RAID set. It allows to allocate free space
on any operational PV already holding parts of the image component pair.
A full search for duplicate PVs in the case of pvs -a
is only necessary when duplicates have previously been
detected in lvmcache. Use a global variable from lvmcache
to indicate that duplicate PVs exist, so we can skip the
search for duplicates when none exist.
When multiple duplicate devices are specified on the
command line, the PV is processed once for each of them,
but pv->dev is the device used each time.
This overrides the PV device to reflect the duplicate
device that was specified on the command line. This is
done by hacking the lvmcache to replace pv->dev with the
device of the duplicate being processed. (It would be
preferable to override pv->dev without munging the content
of the cache, and without sprinkling special cases throughout
the code.)
This override only applies when multiple duplicate devices are
specified on the command line. When only a single duplicate
device of pv->dev is specified, the priority is to display the
cached pv->dev, so pv->dev is not overridden by the named
duplicate device.
In the examples below, loop3 is the cached device referenced
by pv->dev, and is given priority for processing. Only after
loop3 is processed/displayed, will other duplicate devices
loop0/loop1 appear (when requested on the command line.)
With two duplicate devices, loop0 and loop3:
# pvs
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop0
PV VG Fmt Attr PSize PFree
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m
# pvs /dev/loop3
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop0
PV VG Fmt Attr PSize PFree
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m
# pvs /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop0
PV VG Fmt Attr PSize PFree
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m
# pvs -o+dev_size /dev/loop0 /dev/loop3
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop0
PV VG Fmt Attr PSize PFree DevSize
/dev/loop0 loopa lvm2 a-- 12.00m 12.00m 16.00m
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
With three duplicate devices, loop0, loop1, loop3:
# pvs -o+dev_size
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop3
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop1
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop3 /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop0 loopa lvm2 a-- 12.00m 12.00m 16.00m
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop3 /dev/loop1
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop1 loopa lvm2 a-- 12.00m 12.00m 32.00m
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop0 /dev/loop1
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop1 loopa lvm2 a-- 12.00m 12.00m 32.00m
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
# pvs -o+dev_size /dev/loop0 /dev/loop1 /dev/loop3
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop1 not /dev/loop0
Found duplicate PV XhLbpVo0hmuwrMQLjfxuAvPFUFZqD4vr: using /dev/loop3 not /dev/loop1
PV VG Fmt Attr PSize PFree DevSize
/dev/loop0 loopa lvm2 a-- 12.00m 12.00m 16.00m
/dev/loop1 loopa lvm2 a-- 12.00m 12.00m 32.00m
/dev/loop3 loopa lvm2 a-- 12.00m 12.00m 32.00m
Processes a PV once for each time a device with its PV ID
exists on the command line.
This fixes a regression in the case where:
. devices /dev/sdA and /dev/sdB where clones (same PV ID)
. the cached VG references /dev/sdA
. before the regression, the command: pvs /dev/sdB
would display the cached device clone /dev/sdA
. after the regression, pvs /dev/sdB would display nothing,
causing vgimportclone /dev/sdB to fail.
. with this fix, pvs /dev/sdB displays /dev/sdA
Also, pvs /dev/sdA /dev/sdB will report two lines, one for each
device on the command line, but /dev/sdA is displayed for each.
This only works without lvmetad.
Support error_if_no_space feature for thin pools.
Report more info about thinpool status:
(out_of_data (D), metadata_read_only (M), failed (F) also as health
attribute.)
API for seg reporting is breaking internal lvm coding - it cannot
use vgmem mem pool for allocation of reported value.
So use separate pool instead of 'vgmem' for non vg related allocations
Add consts for many function params - but still many other are left
for now as non-const - needs deeper level of change even on libdm side.
An 'lvconvert --repair $RAID_LV" to replace a failed leg of a multi-segment
RAID10/4/5/6 logical volume can lead to allocation of (parts of) the replacement
image component pair on the physical volume of another image component
(e.g. image 0 allocated on the same PV as image 1 silently impeding resilience).
Patch fixes this severe resilince issue by prohibiting allocation on PVs
already holding other legs of the RAID set. It allows to allocate free space
on any operational PV already holding parts of the image component pair.
Normally, if there are partitions defined on top of device-mapper
device, there should be a device-mapper device created for each
partiton on top of the old one and once the underlying DM device
is used by another devices (partition mappings in this case),
it can't be used as a PV anymore.
However, sometimes, it may happen the partition mappings are
missing - either the partitioning tool is not creating them if
it does not contain full support for device-mapper devices or
the mappings were removed.
Better safe than sorry - check for partition header on DM devs
and filter them out as unsuitable for PVs in case the check is
positive. Whatever the user is doing, let's do our best to prevent
unwanted corruption (...by running pvcreate on top of such device
that would corrupt the partition header).
Better than previous patch which changed log_warn to log_error -
we can have multiple MDAs and if one of them fails to be written,
we can still continue with other MDAs if we're in a mode where
we can handle missing PVs - so keep the log_warn for single
failed MDA write as it was before.
However, add log_error with "Failed to write VG <vg_name>." in
case we're not handling missing PVs or no MDA was written at all
during VG write process. This also prevents an internal error in
which the vg_write fails and we're not issuing any other log_error
in vg_write caller or above, so we end up with:
"Internal error: Failed command did not use log_error".
At first, all snapshot-origins where marked as unusable unconditionally
here, but we can't cut off whole snapshot-origin use in a stack just
because of this possible mirror state. This whole "device_is_usable"
check was even incorrectly part of persistent filter before commit
a843d0d97c66aae1872c05b0f6cf4bda176aae2 (where filter cleanup was
done).
The persistent filter is used only if obtain_device_list_from_udev=0,
which means that the former check for snapshot-origin here had not even
been hit with default configuration for a few years before commit
a843d0d97c66aae1872c05b0f6cf4bda176aae2 (the check for snapshot-origin and
skipping of this LV was introduced with commit a71d6051ed
back in 2010).
The obtain_device_list_from_udev=1 (and hence not using persistent
filter and hence not hitting this check for snapshot-origins and skipping) has been
in action since commit edcda01a1e (that is 2011).
So for 3 years this condition was not even checked with default configuration,
making it superfluous.
This all changed in 2014 with commit 8a843d0d97
where "filter-usable" is introduced and since then all snapshot-origins
have been marked as unusable more often than before and making snapshot-origins
practically unusable in a stack.
This patch removes this incorrect check from commit a71d6051ed
which caused snapshot-origins to be unusable more often recently.
If we want to fix this eventually in a correct way, we need to look
down the stack and if snapshot-origin is hit and there's a blocked
mirror underneath, only then mark the device as unusable. But mirrors
in stack are not supported anymore so it's questionable whether it's
worth spending more time on this at all...
$ lvcreate -l1 -m1 --type mirror vg
Logical volume "lvol0" created.
$ lvconvert --type raid1 vg/lvol0
Before:
$ lvs -a vg
LV VG Active Attr LSize Cpy%Sync Layout Role
lvol0 vg active rwi-a-r--- 4.00m 100.00 raid,raid1 public
[lvol0_mimage_0_rimage_0] vg active iwi-aor--- 4.00m linear private,raid,image
[lvol0_mimage_1_rimage_1] vg active iwi-aor--- 4.00m linear private,raid,image
[lvol0_rmeta_0] vg active ewi-aor--- 4.00m linear private,raid,metadata
[lvol0_rmeta_1] vg active ewi-aor--- 4.00m linear private,raid,metadata
Incorrect name: lvol0_mimage_0_rimage_0
With this patch applied:
$ lvs -a vg
LV VG Active Attr LSize Cpy%Sync Layout Role
lvol0 vg active rwi-a-r--- 4.00m 100.00 raid,raid1 public
[lvol0_rimage_0] vg active iwi-aor--- 4.00m linear private,raid,image
[lvol0_rimage_1] vg active iwi-aor--- 4.00m linear private,raid,image
[lvol0_rmeta_0] vg active ewi-aor--- 4.00m linear private,raid,metadata
[lvol0_rmeta_1] vg active ewi-aor--- 4.00m linear private,raid,metadata
Proper name: lvol0_rimage_0
When mirror has missing PVs and there are mirror images on those missing
PVs, we delete the images and during this delete operation, we also
reactivate the LV. But if we're trying to reactivate the LV in cluster
which is not active and at the same time cmirrord is not running (which
is OK since we may have created the mirror LV as inactive), we end up
with:
"Error locking on node <node_name>: Shared cluster mirrors are not available."
That is because we're trying to activate the mirror LV without cmirrord.
However, there's no need to do this reactivation if the mirror LV (and
hence it's sub LVs) were not activated before.
This issue caused failure in mirror-vgreduce-removemissing.sh test
recently with this sequence (excerpt from the test script):
prepare_lvs_
lvcreate -an -Zn -l2 --type mirror -m1 --nosync -n $lv1 $vg "$dev1" $dev2" "$dev3":$BLOCKS
mimages_are_on_ $lv1 "$dev1" "$dev2"
mirrorlog_is_on_ $lv1 "$dev3"
aux disable_dev "$dev2"
vgreduce --removemissing --force $vg
The important thing about that test is that we're not running cmirrord,
we're activating the mirror with "-an" so it's inactive and then
vgreduce --removemissing tries to reactivate the mirror images
as part of the _delete_lv function call inside and since cmirrord
is not running, we end up with the "Shared cluster mirrors are not
available." error.
When creating cluster mirrors while they're not supposed to be activated
immediately after creation, we don't need to check for cmirrord availability.
We can just create these mirrors and let the check to be done on activation
later on. This is addendum for commit cba6186325.
When creating/activating clustered mirrors, we should have cmirrord
available and running. If it's not, we ended up with rather cryptic
errors like:
$ lvcreate -l1 -m1 --type mirror vg
Error locking on node 1: device-mapper: reload ioctl on failed: Invalid argument
Failed to activate new LV.
$ vgchange -ay vg
Error locking on node node 1: device-mapper: reload ioctl on failed: Invalid argument
This patch adds check for cmirror availability and it errors out
properly, also giving a more precise error messge so users are able
to identify the source of the problem easily:
$ lvcreate -l1 -m1 --type mirror vg
Shared cluster mirrors are not available.
$ vgchange -ay vg
Error locking on node 1: Shared cluster mirrors are not available.
Exclusively activated cluster mirror LVs are OK even without cmirrord:
$ vgchange -aey vg
1 logical volume(s) in volume group "vg" now active
Since GET_FIELD_RESERVED_VALUE always returns a pointer, don't reference
it with "&" when used - we already have that pointer value (this is an
addendum to recent commit 028ff30947).
Only GET_TYPE_RESERVED_VALUE needs to be referenced with "&" as it
returns directly the value of that type.
We have to use empty list, not NULL if we want to denote that the list
has no items. Otherwise, the code further can segfault as it expects
there's always a sane value (= some list), including empty list,
but never NULL.
Use helper macros to handle reserved values and also define "undefined"
reserved value as:
FIELD_RESERVED_VALUE(cache_policy, cache_policy_undef, "", "", "undefined")
Which means:
- print "" if the cache_policy value is undefined (the first name for this reserved value is "")
- recognize "undefined" reserved name as synonym to ""
(so statements like "lvs -S cache_policy=undefined" are still recognized)
Avoid making a copy of the keyword which is already registered in
values.h for "unmanaged" (vg_mda_copies field) and "auto" reserved
value (lv_read_ahead field). Also use helper macros to handle these
reserved - this is the correct approach - just do not copy the same
thing again and do not mix it! The GET_FIELD_RESERVED_VALUE and
GET_FIRST_RESERVED_NAME macros guarantees this - use it!
In addition to that, rename reserved values:
vg_mda_copies --> vg_mda_copies_unmanaged
lv_read_ahead --> lv_read_ahead_auto
So the field reserved values follows this scheme:
"<field_name>_<reserved_value_name>".
The same applies for type reserved values with this scheme:
"<report type name in lowercase>_<reserved_value_name>"
Add a comment about this scheme for others to follow as well
when adding new fields and their reserved values. This makes
it a bit easier to read the code then.
RESERVED(id) --> GET_TYPE_RESERVED_VALUE(id)
FIRST_NAME(id) --> GET_FIRST_RESERVED_NAME(id)
Also add GET_FIELD_RESERVED_VALUE(id) macro to get per-field reserved value.
This makes it much more readable and hopefully it'll make it
easier to use these helper macros when adding new reporting
fields with reserved values if needed.
The cache policy name taken as LV segment property must be duped
for report as the VG/LV/seg structure is destroyed after processing,
reporting happens later:
$ valgrind lvs -o+cache_policy
...
==16589== Invalid read of size 1
==16589== at 0x54ABCC3: dm_report_compact_fields
(libdm-report.c:1739)
==16589== by 0x153FC7: _report (reporter.c:619)
==16589== by 0x1540A6: lvs (reporter.c:641)
==16589== by 0x148021: lvm_run_command (lvmcmdline.c:1452)
==16589== by 0x1495CB: lvm2_main (lvmcmdline.c:1907)
==16589== by 0x164712: main (lvm.c:21)
==16589== Address 0x7d465f2 is 8,338 bytes inside a block of size
16,384 free'd
==16589== at 0x4C2ACE9: free (in
/usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==16589== by 0x54B8C85: _free_chunk (pool-fast.c:318)
==16589== by 0x54B84FB: dm_pool_destroy (pool-fast.c:78)
==16589== by 0x1E59C7: _free_vg (vg.c:78)
==16589== by 0x1E5A6D: release_vg (vg.c:95)
==16589== by 0x159B6E: _process_lv_vgnameid_list (toollib.c:1967)
==16589== by 0x159DD7: process_each_lv (toollib.c:2030)
==16589== by 0x153ED8: _report (reporter.c:598)
==16589== by 0x1540A6: lvs (reporter.c:641)
==16589== by 0x148021: lvm_run_command (lvmcmdline.c:1452)
==16589== by 0x1495CB: lvm2_main (lvmcmdline.c:1907)
==16589== by 0x164712: main (lvm.c:21)
When we split leg from raid - we take a proper new lock for a new LV.
However for now activation checks only 'existince' of device UUID,
but it's not validating device has a proper name.
As a quick fix call suspend()/resume() to rename after split mirror.
