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.
Zdenek Kabelac [Wed, 18 Mar 2015 15:19:58 +0000 (16:19 +0100)]
pvscan: check sysfs dev entry before rescan
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.
Zdenek Kabelac [Tue, 17 Mar 2015 08:44:26 +0000 (09:44 +0100)]
lvchange: disable persistent minors for pools
There is no reason to support persistent major/minor numbers
for pool volumes - it's only meant to be supported for filesystems
(since i.e. nfs may need to keep volume on a persistent device node.)
Support for pools is now explicitely disabled and documented.
Peter Rajnoha [Mon, 16 Mar 2015 16:00:36 +0000 (17:00 +0100)]
lvmetad: pvscan: do not scan and read ignored MDAs from PVs
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to cache out-of-date metadata in case other PVs with fresh metadata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).
The iscsi-shutdown.service is the one responsible for logging out
iscsi sessions so blk-availability.service (running the blkdeactivate
script) should be run before that on shutdown (so we need to use
After=iscsi-shutdown.service because "After" relates to starting
the service and the opposite order is automatically applied on
stopping the service at shutdown).
David Teigland [Tue, 10 Mar 2015 18:56:25 +0000 (13:56 -0500)]
reporting: should not fail with foreign option if lvm1 pvs exist
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.
Zdenek Kabelac [Mon, 9 Mar 2015 13:16:41 +0000 (14:16 +0100)]
display: fix return values
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).
David Teigland [Mon, 9 Mar 2015 18:27:34 +0000 (13:27 -0500)]
system_id: detect an lvm1 system id
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.
David Teigland [Fri, 6 Mar 2015 15:55:24 +0000 (09:55 -0600)]
tests: fix systemid uname test
The string reported by uname -n may include characters
that lvm omits from the system id (like parens, as seen
on a test machine.) Check against the final system id
string that lvm uses.
Zdenek Kabelac [Fri, 6 Mar 2015 09:22:33 +0000 (10:22 +0100)]
lib: drop unneeded vg_read call
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'.
Zdenek Kabelac [Fri, 6 Mar 2015 09:24:26 +0000 (10:24 +0100)]
lib: reduce parsing in vgname_from_mda
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.
Zdenek Kabelac [Mon, 26 Mar 2012 11:35:26 +0000 (13:35 +0200)]
lib: avoid reparsing same metadata
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....)
David Teigland [Wed, 4 Mar 2015 17:30:53 +0000 (11:30 -0600)]
system_id: make new VGs read-only for old lvm versions
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.
Peter Rajnoha [Thu, 5 Mar 2015 09:45:29 +0000 (10:45 +0100)]
report: fix seg_monitor field to display monitoring status for thick snapshots and mirrors
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
Peter Rajnoha [Wed, 4 Mar 2015 13:08:47 +0000 (14:08 +0100)]
config: add CFG_DISABLED flag and mark system_id settings with that flag
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.
Peter Rajnoha [Wed, 4 Mar 2015 09:44:31 +0000 (10:44 +0100)]
vgremove: select: direct selection to be done per-VG, not per-LV
Though vgremove operates per VG by definition, internally, it
actually means iterating over each LV it contains to do the
remove.
So we need to direct selection a bit in this case so that the
selection is done per-VG, not per-LV.
That means, use processing handle with void_handle.internal_report_for_select=0
for the process_each_lv_in_vg that is called later in vgremove_single fn.
We need to disable internal selection for process_each_lv_in_vg
here as selection is already done by process_each_vg which calls
vgremove_single. Otherwise selection would be done per-LV and not
per-VG as we intend!
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.
Jose Castillo [Fri, 27 Feb 2015 14:38:34 +0000 (15:38 +0100)]
initscripts: lvm2-monitor: implement status action
Two new functions added in the init script: rh_status and rh_status_q.
First one to be used in status() and second one to be used in start(),
stop(), force_stop(). Check for 'dmeventd' added and print list of
lvs being monitored in status().
Check the simple condition first (variable), then the function return value
(which in this case certainly takes more time to evaluate) - save some time.
Peter Rajnoha [Fri, 27 Feb 2015 12:32:52 +0000 (13:32 +0100)]
report: fix handling of reports with pure label fields
Two problems fixed by this patch:
- PV tags were not recognized at all when using them with pvs
report that has only label fields (regression since 2.02.105)
- incorrect persistent .cache file to be generated after pvs
report that has only label fields (regression since 2.02.106)
David Teigland [Thu, 26 Feb 2015 22:06:32 +0000 (16:06 -0600)]
system_id: remove unwanted foreign error for some commands
Commands that can never use foreign VGs begin with
cmd->error_foreign_vgs = 1. This tells the vg_read
lib layer to print an error as soon as a foreign VG
is read.
The toollib process_each layer also prints an error if a
foreign VG is read, but is more selective about it. It
won't print an error if the command did not explicitly
name the foreign VG. We want to silently ignore foreign VGs
unless a command attempts to use one explicitly.
So, foreign VG errors are printed from two different layers:
vg_read (lower layer) and process_each (upper layer).
Commands that use toollib process_each, only want errors from
the process_each layer, not from both layers. So, process_each
disables the lower layer vg_read error message by setting
error_foreign_vgs = 0.
Commands that do not use toollib process_each, want errors
from the vg_read layer, otherwise they would get no error
message. The original cmd->error_foreign_vgs setting
enables this error.
(Commands that are allowed to operate on foreign VGs always
begin with cmd->error_foreign_vgs = 0, and all the commands
in this group use toollib process_each with the selective
error reporting.)
David Teigland [Wed, 25 Feb 2015 17:33:11 +0000 (11:33 -0600)]
lvchange: deactivate is always possible in foreign vgs
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.
Jonathan Brassow [Wed, 25 Feb 2015 20:42:15 +0000 (14:42 -0600)]
cmirror: Adjust region size to work around CPG msg limit to avoid hang.
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.