1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-22 17:35:59 +03:00
Commit Graph

696 Commits

Author SHA1 Message Date
Alasdair G Kergon
2872e8c289 alloc: Add A_PARTITION_BY_TAGS to avoid sharing.
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.)
2015-04-10 21:57:52 +01:00
Alasdair G Kergon
f1e3e99169 alloc: Log PV tags when reserving areas. 2015-03-26 21:13:26 +00:00
Alasdair G Kergon
e8fa3354f0 alloc: Pass alloc_handle through to _reserve_area. 2015-03-26 20:32:59 +00:00
Alasdair G Kergon
f9d74ba3d1 alloc: Only report cling tag errors once. 2015-03-26 19:43:51 +00:00
Alasdair G Kergon
4b1219ee87 metadata: Move alloc_handle init/destroy fns. 2015-03-26 18:44:24 +00:00
Peter Rajnoha
8759f7d755 metadata: vg: add removed_lvs field to collect LVs which have been removed
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.
2015-03-24 08:43:08 +01:00
Peter Rajnoha
c9f021de0b metadata: process_each_lv_in_vg: get the list of LVs to process first, then do the processing
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).
2015-03-24 08:43:07 +01:00
Jonathan Brassow
dd0ee35378 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.
2015-02-25 14:42:15 -06:00
Zdenek Kabelac
a18d789684 cleanup: simplify error path code
Mempool needs to free only with first alllocated element,
everything allocated afterwards is released as well.
2015-02-19 14:44:04 +01:00
Zdenek Kabelac
4c184e9d6b cleanup: drop unused value assign
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.
2015-02-19 14:43:25 +01:00
Peter Rajnoha
6b4066585f filters: no need to refresh filters/rescan if no signature is wiped during pvcreate at all
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.
2015-02-17 09:46:34 +01:00
Zdenek Kabelac
2e35c68122 lv_manip: add for_each_sub_lv_except_pools()
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.
2015-01-30 12:33:52 +01:00
Zdenek Kabelac
2908ab3eed thin: errrorwhenfull support
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.)
2015-01-14 14:52:05 +01:00
Peter Rajnoha
509650ec4c cmirror: do not check for cmirror availability when creating deactivated cluster mirrors
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.
2015-01-06 09:59:04 +01:00
Peter Rajnoha
cba6186325 cmirror: check for cmirror availability during cluster mirror creation and activation
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
2015-01-05 16:54:07 +01:00
Alasdair G Kergon
a057f40155 mirror: Validate raid region size config setting.
If necessary, round down to a power of 2 the raid/mirror region size
taken from the config files.
2014-12-03 22:47:08 +00:00
Petr Rockai
2c3db52356 metadata: Add cache_policy to lvcreate_params and honour it. 2014-11-27 20:20:48 +01:00
Zdenek Kabelac
d7985ebead thin: fix error path
Print pool name and not the origin name.
2014-11-19 18:58:30 +01:00
Zdenek Kabelac
8cb79dad0b pool: fix removal of pool metadata spare
Since we support device stack of pools over pool
(thin-pool with cache data volume) the existing code
is no longer able to detect orphan _pmspare.

So instead do a _pmspare check after volume removal,
and remove spare afterwards.
2014-11-13 13:09:07 +01:00
Alasdair G Kergon
9a5910bdf9 pre-release 2014-11-11 14:13:00 +00:00
Zdenek Kabelac
8121074fda cache: pending_delete fixes 2014-11-11 13:32:41 +01:00
Zdenek Kabelac
a7fc108298 mirror: layer remove doesn't work properly with mirrors 2014-11-10 22:32:43 +01:00
Zdenek Kabelac
e5d3f81285 cleanup: indents comments backtraces 2014-11-10 22:05:49 +01:00
Zdenek Kabelac
f5e265a07f cache: use LV_PENDING_DELETE 2014-11-10 22:05:49 +01:00
Zdenek Kabelac
6a41286c01 snapshot: virtual save commit
More efficient spare volume creation. Save 1 extra commit
and properly activate this volume according to our cluster
activation rules (using lv_active_change() for this).
2014-11-10 22:05:48 +01:00
Zdenek Kabelac
ff30783a4f lv_manip: remove_layer_from_lv restore lv name
Since we 'layer' for cache origin which and we support dropping
cache layer - we need to restore origin name in case
the origin LV is more complex target  - i.e. raid.

Drop _corig from name

Cleanup and rename parent -> parent_lv.
2014-11-10 22:05:48 +01:00
Zdenek Kabelac
979d12eb4a lvremove: few more validation before archive
Before we finally call archive, we still can do a few more validation
without modifying VG. So delay archive() call few lines.
2014-11-10 22:05:48 +01:00
Zdenek Kabelac
09e32e4add pools: prompt only when metadata lv exists
If the VG has no pool metadata volume, don't prompt about
removal of _pmspare.
2014-11-10 21:35:56 +01:00
Zdenek Kabelac
138c2417c0 pools: after fail of vg_write no restore
Revert part of commit 51a29e6056,
it's probably bad idea to continue with any recovery, when
vg_write() or vg_commit() fail - so it's better to leave it as it is.
2014-11-10 21:34:02 +01:00
Zdenek Kabelac
94e4d5fcf5 cleanup: avoid dm_list size calc in common path
Calculate dm_list_size only when there is not just a single
ont segment in list - so it's only counted on error path.
2014-11-05 15:30:59 +01:00
Zdenek Kabelac
a3a2c792a8 lv: lv_active_change add needs_exclusive flag
Let's use this function for more activations in the code.

'needs_exlusive' will enforce exlusive type for any given LV.

We may want to activate LV in exlusive mode, even when we know
the LV (as is) supports non-exlusive activation as well.

lvcreate -ay   -> exclusive & local
lvcreate -aay  -> exclusive & local
lvcreate -aly  -> exclusive & local
lvcreate -aey  -> exclusive  (might be on any node).
2014-11-05 15:29:43 +01:00
Zdenek Kabelac
a82dcadf66 snapshot: no snapshot of any cache type LVs
Unsupported as of now.
2014-11-05 15:28:38 +01:00
Zdenek Kabelac
8563c3e1a9 thin: check for new pool before creating thin volume
Call check_new_thin_pool() to detect in-use thin-pool.
Save extra reactivation of thin-pool when thin pool is not active.
(it's now a bit more expensive to invoke thin_check for new pools.)

For new pools:

We now active locally exclusively thin-pool as 'public' LV.
Validate transaction_id is till 0.
Deactive.
Prepare create message for thin-pool and exclusively active pool.
Active new thin LV.
And deactivate thin pool if it used to be inactive.
2014-11-04 15:29:19 +01:00
Zdenek Kabelac
b7bc28b7b7 cache: wipe cache-pool before reuse
Before we reuse cache-pool - we need to ensure metadata volume
has wiped header.
2014-11-03 14:19:33 +01:00
Zdenek Kabelac
cb931eed72 cleanup: rename virtual_extents
Use standard 'virtual_extents' naming.
Move virtual_size into  'lcp' struct out of lvcreate_params.
2014-11-03 14:19:33 +01:00
Zdenek Kabelac
9a6cf3d9f4 cleanup: prints and stacks 2014-10-30 23:58:49 +01:00
Zdenek Kabelac
0da749c9a4 cleanup: remove unneeded assign 2014-10-30 23:58:49 +01:00
Zdenek Kabelac
205e3ff888 lvcreate: delay check for free extents
As the rounding for cache creation may change the value of extents
postpone check for free extents.
2014-10-26 18:37:13 +01:00
Zdenek Kabelac
c9fbbf48ba cache: support more args
Unlike with thin-pool - with cache we support all args also
directly when create cache volume.

So the result of 'separate' cache-pool creation and setting its
options should give same result as specifying those args
during cache creation.

Cache-pool values are used as defaults if the params are
not specified with cache creation.
2014-10-26 18:37:13 +01:00
Zdenek Kabelac
ff2e8b0de6 thin: simplify thin volume creation
Move code for creation of thin volume into a single place
out of lv_extend().  This allows to drop extra pool arg
for alloc_lv_segment() && lv_extend() and makes code
more easier to read and follow.
2014-10-26 18:37:13 +01:00
Zdenek Kabelac
52dfa6dd44 cache: apply chunk rounding also for cache creation.
When we create volumes with chunk size bigger then extent size
we try to round up to some nearest chunk boundary.
Until now we did this for thins - use same logic for
cache volumes.
2014-10-26 18:37:13 +01:00
Zdenek Kabelac
25307e4add pools: prefetch pool_lv and origin_lv
Load pool_lv and origin_lv at a single place and
use it for more types as they need them.
2014-10-26 18:37:13 +01:00
Zdenek Kabelac
fd00d2c6ec cache: separate func for pool/origin validation
Split validate_lv_cache_create() into 2 function
once is called for pool validation other for origin.

It's less misleading the accepting NULLs.
2014-10-24 16:39:32 +02:00
Zdenek Kabelac
d13239b054 lvcreate: new validation code
Refactor lvcreate code.

Prefer to use arg_outside_list_is_set() so we get automatic 'white-list'
validation of supported options with different segment types.

Drop used lp->cache, lp->cache and use seg_is_cache(), seg_is_thin()

Draw clear border where is the last moment we could change create
segment type.

When segment type is given with --type - do not allow it to be changed
later.

Put together tests related to individual segment types.

Finish cache conversion at proper part of lv_manip code after
the vg_metadata are written - so we could correcly clean-up created
stripe LV for cache volume.
2014-10-24 16:39:32 +02:00
Zdenek Kabelac
51a29e6056 pools: error path removes metadata
When pool volume is create - it has already managed to update metadata
once within lv_extend - so revert new created LV here.
2014-10-24 16:39:32 +02:00
Zdenek Kabelac
21798418fc cleanup: better stack tracing
Show proper backtrace on error path.
2014-10-24 16:39:32 +02:00
Zdenek Kabelac
37bd12e7ba cleanup: lv_manip drops duplicated internal error
No reason to check the same condition twice with internal error.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
f2ea890423 cleanup: don't backup twice 2014-10-24 16:39:31 +02:00
Zdenek Kabelac
a4788b4994 cleanup: move code for checking vg features
Move code that checks for needed VG features into separate
smaller function.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
b726003b85 cleanup: simplier lv name construction
Using NAME_LEN max lv name size.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
cf674142a2 cleanup: move code 2014-10-24 16:39:31 +02:00
Zdenek Kabelac
4f8ed25a99 cleanup: indent
Various indents and tiny error message improvements.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
f1cb267274 lv_manip: preserve read-only status in layered volume 2014-10-24 16:39:31 +02:00
Zdenek Kabelac
be3f666ef3 lv_manip: improve should_wipe reporting
We want to print smarter warning message only when
the zeroing was not provided on the first zeroable segment
of newly created LV.

Put warning within _should_wipe_lv function to avoid reevaluation
of same conditions twice.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
8a2f553b7d lvcreate: move tests that requires activation
Tests that needs active kernel driver put in single place.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
f7cfea251c lvcreate: better logging for temporary device
Hide creation of temporary LVs and print them only in verbose mode.

e.g. hides confusing message about creation of _pmspare
device during creation of pool.
2014-10-24 16:39:31 +02:00
Zdenek Kabelac
9411c19b31 segments: introduce lvseg_name
Instead of segtype->ops->name()  introduce lvseg_name().

This also allows us to leave name() function 'empty' for default
return of segtype->name.

TODO: add functions for rest of ops->
2014-10-24 16:39:30 +02:00
Zdenek Kabelac
0cbb381e15 lvcreate: regression fix on lv name limits
When we are given an existing LV name - it needs to be allowed
to pass in even restricted name as the LV could have existed
long before we introduced some new restriction on prefix/suffix.i

Fix the regression on name limits and drop restriction to be applied
on any existing LVs - only the new created LV names have to be
complient with current name restrictions.

FIXME: we are currently using restricted names incorrectly in few
other places - device_is_usable() skips restricted names,
and udev flags are also incorrectly set for restricted names
so these LVs are not getting links properly.
2014-10-07 15:52:31 +02:00
Zdenek Kabelac
27ff61d012 cleanup: rename _name, _lv and _size
Use some standard names for lv char * names, logical_volume * lv
and sizes internally.
2014-10-07 15:51:53 +02:00
Zdenek Kabelac
d46c2f1c94 cache: improve creation code
Move code to better locations.
Improve test and remove invalid ones
(i.e. no reason to require cache size to be >= then origin).

Correctly comment where the code is doing actual conversion
of other existing volume - we do already a similar thing with
external origins.

Lots of new command line options and combinations is now supported.
Hopefully older syntax still works as well.

lvcreate --cache --cachepool vg/pool  -l1
lvcreate --type cache --cachepool vg/pool  -l1
lvcreate --type cache-pool vg/pool  -l1
lvcreate --type cache-pool --name pool vg  -l1
... and many many more ...
2014-10-06 15:18:06 +02:00
Zdenek Kabelac
189d0f8e1d lvremove: detect cluster wide active state
If the LV is active remotely do not skip question for deactivation.
2014-10-06 15:18:06 +02:00
Zdenek Kabelac
75119d85b3 cache: lvremoved cached LV removes cache pool
When we remove cached volume, whole tree must go down.
There is  'lvconvert --splitcache' to preserve cache pool.
2014-10-06 15:18:06 +02:00
Zdenek Kabelac
cb273fe164 pools: move removal of _pmspare
Since _pmspare is internal volume move it to
lv_remove_single -  so it's automatically removed with
last remove thin-pool.

lv_remove_with_dependencies() is not always used for pool removal.
2014-10-06 15:18:06 +02:00
Zdenek Kabelac
0662391ff6 lv_manip: fix out of order backup
We had only 2 remaining places that were leaving unbackuped metadata.
2014-10-06 15:18:05 +02:00
Heinz Mauelshagen
45f57477f4 cleanup: Use segtype.h definitions of segment type names wherever possible
We are not using already defined segement type names where we could.

There is a lot of other places in device-mapper and LVM2 we have those
hardcoded so we should better finally have a common interface in
libdevmapper to avoid this.
2014-09-24 15:24:41 +02:00
Zdenek Kabelac
a8aee7dba2 activate: update lv_check_not_in_use: API
Use of lv_info() internally in lv_check_not_in_use(),
so it always could use with_open_count properly.

Skip sysfs() testing in open_count == 0 case.

Accept just 'lv' pointer like other functions.

The function has 'built-in' lv_is_active_locally check,
which however is not what we need to check in many place.
For now at least remotely active snapshot merge is
detected and for this case merge on next activation is scheduled.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
a84d0d0c7b snapshot: use lv_update_and_reload 2014-09-24 10:51:15 +02:00
Zdenek Kabelac
736f40134b mirror: extend adjusted_mirror_region_size API
We use adjusted_mirror_region_size() in two different contexts.

Either on command line -
 here we do want to inform user about reduction of size.

Or in pvmove activation context -
 here we should only use 'verbose' info.
2014-09-24 10:48:02 +02:00
Zdenek Kabelac
fa6482344c lv_update_and_reload: use proper lock holder
When requesting to reload an LV imrove this API to
automatically reload its lock holding LV as in cluster
only top-level LVs are addressable with lock.
2014-09-24 10:48:02 +02:00
Zdenek Kabelac
f1e9e94a5b wipe_lv: move sync_local_dev_names in front
Synchronize things before checking for locally active volume.
2014-09-19 15:51:30 +02:00
Zdenek Kabelac
d8b775f4eb memlock: drop uneeded lock
Avoid doing buffered operation within memory lock.
2014-09-19 01:13:49 +02:00
Alasdair G Kergon
979be63f25 mirrors: Fix checks for mirror/raid/pvmove LVs.
Try to enforce consistent macro usage along these lines:

lv_is_mirror - mirror that uses the original dm-raid1 implementation
               (segment type "mirror")
lv_is_mirror_type - also includes internal mirror image and log LVs

lv_is_raid - raid volume that uses the new dm-raid implementation
             (segment type "raid")
lv_is_raid_type - also includes internal raid image / log / metadata LVs

lv_is_mirrored - LV is mirrored using either kernel implementation
                 (excludes non-mirror modes like raid5 etc.)

lv_is_pvmove - internal pvmove volume
2014-09-16 00:13:46 +01:00
Alasdair G Kergon
2360ce3551 cleanup: Use lv_is_ macros.
Use lv_is_* macros throughout the code base, introducing
lv_is_pvmove, lv_is_locked, lv_is_converting and lv_is_merging.

lv_is_mirror_type no longer includes pvmove.
2014-09-15 21:33:53 +01:00
Zdenek Kabelac
15ba2afdc2 allocation: use vg memory pool
Looks like forgotten memory allocation related to VG used cmd mem pool.
2014-09-12 13:39:58 +02:00
Zdenek Kabelac
a86d9a3b30 lv_rename: actual fix for snapshot
By my rebasing mistake it's been eliminated from previous patch set.
2014-09-09 20:15:51 +02:00
Zdenek Kabelac
c710f02e01 lv_update_and_reload: replace code sequence
Use lv_update_and_reload() and lv_update_and_reload_origin()
to handle write/suspend/commit/resume sequence.

In few places this properly handle vg_revert() after suspend failure,
and also ensures there is metadata backup after successful vg_commit().
2014-09-09 19:20:09 +02:00
Zdenek Kabelac
aee8611af5 lv_manip: remove vg_revert
vg_commit is supposed to have implicit revert handling.
(however as of now it needs fixes).
2014-09-09 19:15:26 +02:00
Zdenek Kabelac
413fc9d3e6 lv_rename: fix snapshot rename
Fix rename operation for snapshot (cow) LV.
Only the snapshot's origin has the lock and by mistake suspend
and resume has been called for the snapshot LV.
This further made volumes unusable in cluster.

So instead of suspend and resuming list of LVs,
we need to just suspend and resume origin.

As the sequence write/suspend/commit/resume
is widely used in lvm2 code base - move it to
new lv_update_and_reload function.
2014-09-09 19:15:24 +02:00
Zdenek Kabelac
319f67b1ab cleanup: add stacktrace for error path 2014-09-08 22:36:42 +02:00
Alasdair G Kergon
2faf416e0e lvextend: Reinstate --nosync logic for mirrors.
Reinstate the logic for syncing extensions of mirrors created with
--nosync.  (Inadvertently disabled by the approximate allocation
changes.)
2014-08-28 00:40:09 +01:00
Zdenek Kabelac
22bfac5dc2 cache: fix allocation size
Commit 0b3d0e79f6 caused regression
in allocation of cache pool. This patch is restoring corect size
for allocation.
2014-08-27 16:47:14 +02:00
Alasdair G Kergon
8b8d21f873 pre-release 2014-08-26 16:34:14 +01:00
Zdenek Kabelac
25fe716b12 cleanup: indent and stacktrack
Add missing stacktrace on error path
and newline indent.
2014-08-26 14:13:07 +02:00
Zdenek Kabelac
0794a10f91 thin: fix volume_list support
Fixing problem, when user sets volume_list and excludes thin pools
from activation. In this case pool return 'success' for skipped activation.

We need to really check the volume it is actually active to properly
to remove queued pool messages. Otherwise the lvm2 and kernel
metadata started to go async since lvm2 believed, messages were submitted.

Add also better check for threshold when create a new thin volume.
In this case we require local activation of thin pool so we are able
to check pool fullness.
2014-08-26 14:10:18 +02:00
Zdenek Kabelac
1ee5e18a7b thin: more forced ignoring of pool failure
Support also 'vgremove -ff' to properly remove even inactive/broken thin pools.
Update messages to use 'print_unless_silent' for the forced case.
2014-08-26 14:09:04 +02:00
Peter Rajnoha
f4e56b2829 cleanup: consolidate lv_layout and lv_role reporting
This patch makes the keyword combinations found in "lv_layout" and
"lv_role" much more understandable - there were some ambiguities
for some of the combinations which lead to confusion before.

Now, the scheme used is:

LAYOUTS ("how the LV is laid out"):
===================================
[linear] (all segments have number of stripes = 1)

[striped] (all segments have number of stripes > 1)

[linear,striped] (mixed linear and striped)

raid (raid layout always reported together with raid level, raid layout == image + metadata LVs underneath that make up raid LV)
  [raid,raid1]
  [raid,raid10]
  [raid,raid4]
  [raid,raid5] (exact sublayout not specified during creation - default one used - raid5_ls)
    [raid,raid5,raid5_ls]
    [raid,raid5,raid6_rs]
    [raid,raid5,raid5_la]
    [raid,raid5,raid5_ra]
  [raid6,raid] (exact sublayout not specified during creation - default one used - raid6_zr)
    [raid,raid6,raid6_zr]
    [raid,raid6,raid6_nc]
    [raid,raid6,raid6_ns]

[mirror] (mirror layout == log + image LVs underneath that make up mirror LV)

thin (thin layout always reported together with sublayout)
  [thin,sparse] (thin layout == allocated out of thin pool)
  [thin,pool] (thin pool layout == data + metadata volumes underneath that make up thin pool LV, not supposed to be used for direct use!!!)

[cache] (cache layout == allocated out of cache pool in conjunction with cache origin)
  [cache,pool] (cache pool layout == data + metadata volumes underneath that make up cache pool LV, not supposed to be used for direct use!!!)

[virtual] (virtual layout == not hitting disk underneath, currently this layout denotes only 'zero' device used for origin,thickorigin role)

[unknown] (either error state or missing recognition for such layout)

ROLES ("what's the purpose or use of the LV - what is its role"):
=================================================================
- each LV has either of these two roles at least:  [public] (public LV that users may use freely to write their data to)

  [public] (public LV that users may use freely to write their data to)
  [private] (private LV that LVM maintains; not supposed to be directly used by user to write his data to)

- and then some special-purpose roles in addition to that:

  [origin,thickorigin] (origin for thick-style snapshot; "thick" as opposed to "thin")
  [origin,multithickorigin] (there are more than 2 thick-style snapshots for this origin)
  [origin,thinorigin] (origin for thin snapshot)
  [origin,multithinorigin] (there are more than 2 thin snapshots for this origin)
  [origin,extorigin] (external origin for thin snapshot)
  [origin,multiextoriginl (there are more than 2 thin snapshots using this external origin)
  [origin,cacheorigin] (cache origin)

  [snapshot,thicksnapshot] (thick-style snapshot; "thick" as opposed to "thin")
  [snapshot,thinsnapshot] (thin-style snapshot)

  [raid,metadata] (raid metadata LV)
  [raid,image] (raid image LV)

  [mirror,image] (mirror image LV)
  [mirror,log] (mirror log LV)
  [pvmove] (pvmove LV)

  [thin,pool,data] (thin pool data LV)
  [thin,pool,metadata] (thin pool metadata LV)

  [cache,pool,data] (cache pool data LV)
  [cache,pool,metadata] (cache pool metadata LV)

  [pool,spare] (pool spare LV - common role of LV that makes it used for both thin and cache repairs)
2014-08-25 16:14:40 +02:00
Peter Rajnoha
993f8d1b3f refactor: rename 'lv_type' field to 'lv_role'
The 'lv_type' field name was a bit misleading. Better one is 'lv_role'
since this fields describes what's the actual use of the LV currently -
its 'role'.
2014-08-25 16:11:40 +02:00
Alasdair G Kergon
0b3d0e79f6 lvresize: Fix raid/mirror and %PE handling code.
Sort out the lvresize calculation code to handle size changes
specified as physical extents as well as logical extents
and to process mirror resizing and raid extensions correctly.

The 'approx alloc' option was masking the underlying problem.
2014-08-22 01:26:14 +01:00
Zdenek Kabelac
dec39b1a5f lv_manip: check for str_list_dup failure 2014-08-19 14:33:06 +02:00
Peter Rajnoha
84860fd54f lv: remove lv_type_name fn
The lv_type_name function is remnant from old code that reported
only single string for the LV type. LV types are now reported
in a more extended way as keyword list that describe the type
precisely (using lv_layout_and_type fn).

The lv_type_name was used in some error messages to display the
type of the LV so just reinstate the old messages back referencing
the type directly with a string - this is enough for error messages.
They don't need to display the LV type as precisely as it's used
on lvs output (which is optimized for selection anyway).
2014-08-19 14:16:39 +02:00
Peter Rajnoha
aec4d0c939 report: also display "mirror" keyword in lv_layout for mirrored mirror log and "cache" keyword in lv_layout for cached cache pool
$ lvs -a -o name,vg_name,attr,layout,type
  LV                    VG     Attr       Layout     Type
  lvol0                 vg     mwi-a-m--- mirror     mirror
  [lvol0_mimage_0]      vg     iwi-aom--- linear     image,mirror
  [lvol0_mimage_1]      vg     iwi-aom--- linear     image,mirror
  [lvol0_mlog]          vg     mwi-aom--- mirror     log,mirror
  [lvol0_mlog_mimage_0] vg     iwi-aom--- linear     image,mirror
  [lvol0_mlog_mimage_1] vg     iwi-aom--- linear     image,mirror

(lvol0_mlog properly displayed as "mirror" layout for mirrored mirror log)

$ lvs -a -o name,vg_name,attr,layout,type
  LV                  VG     Attr       Layout     Type
  lvol0               vg     Cwi---C--- cache,pool cache,pool
  [lvol0_cdata]       vg     Cwi------- linear     cache,data,pool
  [lvol0_cmeta]       vg     ewi------- linear     cache,metadata,pool
  [lvol1_pmspare]     vg     ewi------- linear     metadata,pool,spare
  lvol2               vg     Cwi---C--- cache,pool cache,pool
  [lvol2_cdata]       vg     Cwi---C--- cache      cache,data,pool
  [lvol2_cdata_corig] vg     owi---C--- linear     cache,origin
  [lvol2_cmeta]       vg     ewi------- linear     cache,metadata,pool

(lvol2_cdata properly displayed as cached cache pool data)
2014-08-19 13:58:32 +02:00
Peter Rajnoha
b806836164 report: also display "mirror" keyword in lv_type for pvmove LV and display "multiple" for external origin used for more than one thin snapshot
$ lvs -a -o name,vg_name,attr,layout,type
  LV        VG     Attr       Layout     Type
  lvol0     vg     -wI-a----- linear     linear
  [pvmove0] vg     p-C-aom--- mirror     mirror,pvmove

(added "mirror" for pvmove LV)

$ lvs -a -o name,vg_name,attr,layout,type
  LV              VG     Attr       Layout     Type
  lvol0           vg     ori------- linear     external,multiple,origin,thin
  [lvol1_pmspare] vg     ewi------- linear     metadata,pool,spare
  lvol2           vg     Vwi-a-tz-- thin       snapshot,thin
  lvol3           vg     Vwi-a-tz-- thin       snapshot,thin
  pool            vg     twi-a-tz-- pool,thin  pool,thin
  [pool_tdata]    vg     Twi-ao---- linear     data,pool,thin
  [pool_tmeta]    vg     ewi-ao---- linear     metadata,pool,thin

(added "multiple" for external origin used for more than one
thin snapshot - lvol0 in the example above)
2014-08-19 09:41:41 +02:00
Peter Rajnoha
90c47a4968 report: fix thin external snapshot identification for lv_layout and lv_type fields
Thin snapshots having external origins missed the "snapshot" keyword for
lv_type field. Also, thin external origins which are thin devices (from
another pool) were not recognized properly.

For example, external origin itself can be either non-thin volume (lvol0
below) or it can be a thin volume from another pool (lvol3 below):

Before this patch:

$ lvs -o name,vg_name,attr,pool_lv,origin,layout,type
  Internal error: Failed to properly detect layout and type for for LV vg/lvol3
  Internal error: Failed to properly detect layout and type for for LV vg/lvol3
  LV    VG     Attr       Pool  Origin Layout     Type
  lvol0 vg     ori-------              linear     external,origin,thin
  lvol2 vg     Vwi-a-tz-- pool  lvol0  thin       thin
  lvol3 vg     ori---tz-- pool         unknown    external,origin,thin,thin
  lvol4 vg     Vwi-a-tz-- pool1 lvol3  thin       thin
  pool  vg     twi-a-tz--              pool,thin  pool,thin
  pool1 vg     twi-a-tz--              pool,thin  pool,thin

- lvol2 as well as lvol4 have missing "snapshot" in type field
- lvol3 has unrecognized layout (should be "thin"), but has double
  "thin" in lv_type which is incorrect
- (also there's double "for" in the internal error message)

With this patch applied:

$ lvs -o name,vg_name,attr,pool_lv,origin,layout,type
  LV    VG     Attr       Pool  Origin Layout     Type
  lvol0 vg     ori-------              linear     external,origin,thin
  lvol2 vg     Vwi-a-tz-- pool  lvol0  thin       snapshot,thin
  lvol3 vg     ori---tz-- pool         thin       external,origin,thin
  lvol4 vg     Vwi-a-tz-- pool1 lvol3  thin       snapshot,thin
  pool  vg     twi-a-tz--              pool,thin  pool,thin
  pool1 vg     twi-a-tz--              pool,thin  pool,thin
2014-08-18 15:58:48 +02:00
Jonathan Brassow
4d45302e25 RAID: Fail RAID4/5/6 creation if PE size is less than STRIPE_SIZE_MIN
The maximum stripe size is equal to the volume group PE size.  If that
size falls below the STRIPE_SIZE_MIN, the creation of RAID 4/5/6 volumes
becomes impossible.  (The kernel will fail to load a RAID 4/5/6 mapping
table with a stripe size less than STRIPE_SIZE_MIN.)  So, we report an
error if it is attempted.

This is very rare because reducing the PE size down that far limits the
size of the PV below that of modern devices.
2014-08-15 21:15:34 -05:00
Peter Rajnoha
8e449ebd63 cleanup: gcc warning
metadata/lv_manip.c:269: warning: declaration of "snapshot_count" shadows a global declaration

There's existing function called "snapshot_count" so rename the
variable to "snap_count".
2014-08-15 15:32:04 +02:00
Peter Rajnoha
e8bbcda2a3 Add lv_layout_and_type fn, lv_layout and lv_type reporting fields.
The lv_layout and lv_type fields together help with LV identification.
We can do basic identification using the lv_attr field which provides
 very condensed view. In contrast to that, the new lv_layout and lv_type
fields provide more detialed information on exact layout and type used
for LVs.

For top-level LVs which are pure types not combined with any
other LV types, the lv_layout value is equal to lv_type value.

For non-top-level LVs which may be combined with other types,
the lv_layout describes the underlying layout used, while the
lv_type describes the use/type/usage of the LV.

These two new fields are both string lists so selection (-S/--select)
criteria can be defined using the list operators easily:
  [] for strict matching
  {} for subset matching.

For example, let's consider this:

$ lvs -a -o name,vg_name,lv_attr,layout,type
  LV                    VG     Attr       Layout       Type
  [lvol1_pmspare]       vg     ewi------- linear       metadata,pool,spare
  pool                  vg     twi-a-tz-- pool,thin    pool,thin
  [pool_tdata]          vg     rwi-aor--- level10,raid data,pool,thin
  [pool_tdata_rimage_0] vg     iwi-aor--- linear       image,raid
  [pool_tdata_rimage_1] vg     iwi-aor--- linear       image,raid
  [pool_tdata_rimage_2] vg     iwi-aor--- linear       image,raid
  [pool_tdata_rimage_3] vg     iwi-aor--- linear       image,raid
  [pool_tdata_rmeta_0]  vg     ewi-aor--- linear       metadata,raid
  [pool_tdata_rmeta_1]  vg     ewi-aor--- linear       metadata,raid
  [pool_tdata_rmeta_2]  vg     ewi-aor--- linear       metadata,raid
  [pool_tdata_rmeta_3]  vg     ewi-aor--- linear       metadata,raid
  [pool_tmeta]          vg     ewi-aor--- level1,raid  metadata,pool,thin
  [pool_tmeta_rimage_0] vg     iwi-aor--- linear       image,raid
  [pool_tmeta_rimage_1] vg     iwi-aor--- linear       image,raid
  [pool_tmeta_rmeta_0]  vg     ewi-aor--- linear       metadata,raid
  [pool_tmeta_rmeta_1]  vg     ewi-aor--- linear       metadata,raid
  thin_snap1            vg     Vwi---tz-k thin         snapshot,thin
  thin_snap2            vg     Vwi---tz-k thin         snapshot,thin
  thin_vol1             vg     Vwi-a-tz-- thin         thin
  thin_vol2             vg     Vwi-a-tz-- thin         multiple,origin,thin

Which is a situation with thin pool, thin volumes and thin snapshots.
We can see internal 'pool_tdata' volume that makes up thin pool has
actually a level10 raid layout and the internal 'pool_tmeta' has
level1 raid layout. Also, we can see that 'thin_snap1' and 'thin_snap2'
are both thin snapshots while 'thin_vol1' is thin origin (having
multiple snapshots).

Such reporting scheme provides much better base for selection criteria
in addition to providing more detailed information, for example:

$ lvs -a -o name,vg_name,lv_attr,layout,type -S 'type=metadata'
LV                   VG   Attr       Layout      Type
[lvol1_pmspare]      vg   ewi------- linear      metadata,pool,spare
[pool_tdata_rmeta_0] vg   ewi-aor--- linear      metadata,raid
[pool_tdata_rmeta_1] vg   ewi-aor--- linear      metadata,raid
[pool_tdata_rmeta_2] vg   ewi-aor--- linear      metadata,raid
[pool_tdata_rmeta_3] vg   ewi-aor--- linear      metadata,raid
[pool_tmeta]         vg   ewi-aor--- level1,raid metadata,pool,thin
[pool_tmeta_rmeta_0] vg   ewi-aor--- linear      metadata,raid
[pool_tmeta_rmeta_1] vg   ewi-aor--- linear      metadata,raid

(selected all LVs which are related to metadata of any type)

lvs -a -o name,vg_name,lv_attr,layout,type -S 'type={metadata,thin}'
LV           VG   Attr       Layout      Type
[pool_tmeta] vg   ewi-aor--- level1,raid metadata,pool,thin

(selected all LVs which hold metadata related to thin)

lvs -a -o name,vg_name,lv_attr,layout,type -S 'type={thin,snapshot}'
LV         VG   Attr       Layout     Type
thin_snap1 vg   Vwi---tz-k thin       snapshot,thin
thin_snap2 vg   Vwi---tz-k thin       snapshot,thin

(selected all LVs which are thin snapshots)

lvs -a -o name,vg_name,lv_attr,layout,type -S 'layout=raid'
LV           VG   Attr       Layout       Type
[pool_tdata] vg   rwi-aor--- level10,raid data,pool,thin
[pool_tmeta] vg   ewi-aor--- level1,raid  metadata,pool,thin

(selected all LVs with raid layout, any raid layout)

lvs -a -o name,vg_name,lv_attr,layout,type -S 'layout={raid,level1}'
  LV           VG   Attr       Layout      Type
  [pool_tmeta] vg   ewi-aor--- level1,raid metadata,pool,thin

(selected all LVs with raid level1 layout exactly)

And so on...
2014-08-15 14:50:38 +02:00
Alasdair G Kergon
c7b9f0ab42 lvresize: Allow approximation with +%FREE.
Make lvresize -l+%FREE support approximate allocation.

Move existing "Reducing/Extending' message to verbose level
and change it to say 'up to' if approximate allocation is being used.

Replace it with a new message that gives the actual old and new size or
says 'unchanged'.
2014-08-01 00:35:43 +01:00
Zdenek Kabelac
894eda4707 thin and cache: unify pool common code
Fix get_pool_params to only read params.
Add poolmetadataspare option to get_pool_params.
Move all profile code into update_pool_params.
Move recalculate code into pool_manip.c
2014-07-22 22:41:38 +02:00
Peter Rajnoha
f6001465ef lv_manip: pool-metadata-spare is just a spare LV, not tightly bound to thin or cache 2014-07-07 17:02:06 +02:00
Peter Rajnoha
6b58647848 lv_manip: add get_lv_type_name/lv_is_linear and lv_is_striped helper fns
The get_lv_type_name helps with translating volume type
to human readable form (can be used in reports or
various messages if needed).

The lv_is_linear and lv_is_striped complete the set of
lv_is_* functions that identify exact volume types.
2014-07-04 15:40:17 +02:00