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

1477 Commits

Author SHA1 Message Date
Tony Asleson
4f26eae40f lvmdbusd: Correct pv_seg_start -> pvseg_start
Code was using the incorrect column names for pvseg_start which only
became apparent when utilizing the JSON.
2016-06-28 12:09:28 -05:00
Tony Asleson
0aadd6b0fb lvmdbusd: Correct command line args for JSON
Use the updated syntax.
2016-06-28 12:09:28 -05:00
Tony Asleson
4b337b20d4 lvmdbusd: Remove sorting in db layer
When using the JSON this does not yield a totally sorted list as we don't
have a complete set of LVs, so remove this sort.
2016-06-28 12:09:28 -05:00
Tony Asleson
e514284c65 lvmdbusd: Add hidden in for json too 2016-06-28 12:09:28 -05:00
Tony Asleson
ec45be9976 lvmdbusd: Initial support for json output 2016-06-28 12:09:28 -05:00
David Teigland
a7c45ddc59 lvmetad: two phase vg_update
Previously, a command sent lvmetad new VG metadata in vg_commit().
In vg_commit(), devices are suspended, so any memory allocation
done by the command while sending to lvmetad, or by lvmetad while
updating its cache could deadlock if memory reclaim was triggered.

Now lvmetad is updated in unlock_vg(), after devices are resumed.
The new method for updating VG metadata in lvmetad is in two phases:

1. In vg_write(), before devices are suspended, the command sends
   lvmetad a short message ("set_vg_info") telling it what the new
   VG seqno will be.  lvmetad sees that the seqno is newer than
   the seqno of its cached VG, so it sets the INVALID flag for the
   cached VG.  If sending the message to lvmetad fails, the command
   fails before the metadata is committed and the change is not made.
   If sending the message succeeds, vg_commit() is called.

2. In unlock_vg(), after devices are resumed, the command sends
   lvmetad the standard vg_update message with the new metadata.
   lvmetad sees that the seqno in the new metadata matches the
   seqno it saved from set_vg_info, and knows it has the latest
   copy, so it clears the INVALID flag for the cached VG.

If a command fails between 1 and 2 (after committing the VG on disk,
but before sending lvmetad the new metadata), the cached VG retains
the INVALID flag in lvmetad.  A subsequent command will read the
cached VG from lvmetad, see the INVALID flag, ignore the cached
copy, read the VG from disk instead, update the lvmetad copy
with the latest copy from disk, (this clears the INVALID flag
in lvmetad), and use the correct VG metadata for the command.

(This INVALID mechanism already existed for use by lvmlockd.)
2016-06-28 02:30:31 +01:00
David Teigland
17ad29ebba lvmlockd: fix coverity report 2016-06-28 02:28:49 +01:00
David Teigland
47a29f6b2e lvmlockd: always try to connect to lvmetad
Reconnect to lvmetad if either the send fails (e.g. lvmetad
was restarted since lvmlockd last connected), or if no
lvmetad connection exists (e.g. lvmetad was started after
lvmlockd so no previous connection existed.)
2016-06-20 14:01:46 -05:00
David Teigland
6ae22125c6 vgcfgrestore: use lvmetad disabled state
Previously, vgcfgrestore would attempt to vg_remove the
existing VG from lvmetad and then vg_update to add the
restored VG.  But, if there was a failure in the command
or with vg_update, the lvmetad cache would be left incorrect.
Now, disable lvmetad before the restore begins, and then
rescan to populate lvmetad from disk after restore has
written the new VG to disk.
2016-06-20 11:19:49 -05:00
Tony Asleson
cdf06044e1 lvmdbusd: Create correct LV object type
We were initially looking to see if an LV was hidden and if it was we were
creating an instance of a LvCommon object to represent it.  Thus if we
had a hidden cache pool for example we were missing the methods and
properties for the cache pool.  However, when we create the object path,
any hidden LVs, regardless of type/functionality will be placed in the
hidden path.
2016-06-10 16:05:03 -05:00
Tony Asleson
b81186e535 lvmdbusd: Add method get_object_path_by_lvm_id
The object manager method get_object_by_lvm_id was used in many cases for
the sole reason of getting the object path for the object.  Instead of
retrieving the object and then calling 'dbus_object_path' on the object, we
are adding a method which returns the object path.
2016-06-10 16:04:43 -05:00
Tony Asleson
b717e8fe1d lvmdbusd: Rename get_object_path_by_lvm_id
Renaming to get_object_path_by_uuid_lvm_id, to clarify that both the uuid
and the lvm id are required.
2016-06-10 15:28:42 -05:00
Tony Asleson
41ebed7077 lvmdbusd: Change print statement to py3 syntax 2016-06-10 15:28:42 -05:00
Tony Asleson
b717c5efae lvmdbusd: Bug fix for missing LV properties
When we are processing the LVs we need to build up dbus objects from least
dependent to most dependent, so that we have information available when
constructing.
2016-06-10 15:28:42 -05:00
Tony Asleson
e04705b305 lvmdbusd: Add data_lv and metadata_lv to cache pool interface
See: https://bugzilla.redhat.com/show_bug.cgi?id=1342558
2016-06-10 15:28:42 -05:00
Tony Asleson
3bf43a65fe lvmdbusd: Add roles property to LV common interface
See: https://bugzilla.redhat.com/show_bug.cgi?id=1342559
2016-06-10 15:28:42 -05:00
David Teigland
ed6ffc7a34 lvmetad: handle update failures
If a command gets stuck during an lvmetad update, lvmetad
will cancel that update after the timeout.  The next command
to check the lvmetad will see that lvmetad needs to be
populated because lvmetad will return token of "none" after
a timed out update (same as when lvmetad is not populated
at all after starting.)

If a command gets an error during an lvmetad update, it
will now just quit and leave its updating token in place.
That update will be cancelled after the timeout.
2016-06-07 10:17:00 -05:00
David Teigland
851ccfccaf lvmetad: remove disabled case for "scan error"
Failures while populating lvmetad will be handling
differently in a subsequent commit.
2016-06-07 10:17:00 -05:00
David Teigland
0e7f352c70 lvmetad: define special update in progress string 2016-06-07 10:17:00 -05:00
Peter Rajnoha
fc37ee63c0 make: add generated parts of lvmdbusd to DISTCLEAN_TARGETS for make distclean 2016-06-02 11:58:54 +02:00
Peter Rajnoha
3d333e5a29 coverity: fix warnings about missing return value check for sscanf
All the variables for sscanf in lvmlockctl.c and lvmlockd-sanlock.c are
zeroed before sscanf call so the failure is detected by seeing the zero
value instead of proper one in subsequent code - so use (void) for
sscanf calls to ignore return value here.
2016-05-31 10:06:25 +02:00
Peter Rajnoha
48877e215d coverity: missing check for id_write_format return value 2016-05-31 09:56:10 +02:00
David Teigland
e0c22df5c4 lvmetad: remove old thread locking
The old thread locking is very fine grained and complex,
and appears to have no concurrency advantage.
2016-05-16 14:38:25 -05:00
David Teigland
9f72a52302 lvmetad: add request level thread locking
Use an rwlock for requests that read/write cache state.
Use a mutex for requests that read/write global info.
2016-05-16 14:38:04 -05:00
David Teigland
b1ea27b1e2 lvmetad: disable if device scan fails
If a command begins repopulating the lvmetad cache,
and fails part way through, it should set the disabled
state in lvmetad so other commands don't use bad data.
If a subsequent scan succeeds, the disabled state is
cleared.
2016-05-06 09:00:00 -05:00
David Teigland
3d2fbfe243 lvmetad: set disabled flag when duplicate PVs are seen
When duplicate PVs are detected, set the disabled
flag so that commands will disable use of lvmetad.

This duplicate detection is done by lvmetad itself
when it's told about a single new PV with a PVID
that matches an existing PV on another device.
(This is different from the case where the command
is scanning all devices and detects the duplicate.)

Remove the "altdev" logic that attempted to keep
track of multiple devices for a single PV.  It
is no longer used since lvmetad is disabled in
this case.
2016-05-06 08:59:59 -05:00
Zdenek Kabelac
a88828c845 coverity: check for sys error in clock call 2016-04-22 14:15:36 +02:00
Zdenek Kabelac
d12ba022c8 coverity: keep expression using 64bit
Use 64bit input to remain with whole evaluation in 64bit.
2016-04-22 14:15:36 +02:00
David Teigland
593900b795 lvmetad: use defines for disabled reason strings 2016-04-19 11:45:24 -05:00
David Teigland
a4ef8fa25e lvmetad: add disabled state
A global flag in lvmetad indicates it has been disabled.
Other flags indicate the reason it was disabled.
These flags can be queried using get_global_info.

The lvmetactl debugging utility can set and clear the
disabled flag in lvmetad.  Nothing else sets the
disabled flag yet.

Commands will check these flags after connecting to
lvmetad.  If the disabled flag is set, the command
will not use the lvmetad cache, but revert to disk
scanning.

To test this feature:

$ lvmetactl get_global_info
response = "OK"
global_invalid = 0
global_disable = 0
disable_reason = "none"
token = "filter:3041577944"

$ vgs
(should report VGs from lvmetad)

$ lvmetactl set_global_disable 1

$ lvmetactl get_global_info
response = "OK"
global_invalid = 0
global_disable = 1
disable_reason = "DIRECT"
token = "filter:3041577944"

$ vgs
  WARNING: Not using lvmetad because the disable flag was set directly.
(should report VGs without contacting lvmetad)

$ lvmetactl set_global_disable 0

$ vgs
(should report VGs from lvmetad)
2016-04-19 09:41:18 -05:00
David Teigland
56c68b3476 lvmetad: preemptively check and rescan in commands
Move checking the lvmetad state, and the possible rescan,
out of lvmetad_send() to the start of the command.

Previously, the token mismatch and rescan would occur
within lvmetad_send() for some other request.  Now,
the token mismatch is detected earlier, so the
rescan can be done before the main command is in
progress.  Rescanning deep within the processing of
another command will disturb the lvmcache state of
that other command.

A rescan already exists at the start of the command
for the case where foreign VGs are going to be read.
This same rescan is now also performed when there is
an lvmetad token mismatch (from a changed global_filter).

The commands pvscan/vgscan/lvscan/vgimport are excluded
from this preemptive checking/rescanning for lvmetad
because they want to do rescanning themselves explicitly.

If rescanning devices fails, then lvmetad has not been
correctly repopulated and should not be used, so make
the command revert to not using lvmetad.
2016-04-13 14:05:42 -05:00
Zdenek Kabelac
1be74cfd7f cleanup: gcc warn about comparing int with uint 2016-04-12 11:47:51 +02:00
Zdenek Kabelac
af148a9d77 cleanup: avoid gcc warns
Some older systems may had global declaration to cause gcc warning.
Rename for cases we don't care...
2016-04-08 20:20:16 +02:00
Tony Asleson
d243db5466 lvmdbusd: Add '-' to allowable PV device path chars 2016-03-24 12:27:02 -05:00
Alasdair G Kergon
1216efdf15 activate: Use macros for target and module names. 2016-03-22 17:46:15 +00:00
Tony Asleson
16b1272597 lvmdbusd: Change print statements to log_error
Should not be using print() in the service code.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-21 17:28:51 -05:00
Tony Asleson
ca28ea88be lvmdbusd: Add lvm flight recorder
To help out with debug, when an exception is thrown in the dbus service we
will dump all the information we have on the last 16 commands that were
executed along with the stack strace.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-21 17:28:51 -05:00
Tony Asleson
a0c7875c54 lvmdbusd: Allow PV devices to be referenced by symlink(s)
See: https://bugzilla.redhat.com/show_bug.cgi?id=1318754

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-21 17:28:51 -05:00
Tony Asleson
6d19c14c28 lvmdbus: Fix deprecated warnings for GObject use
While running on F24 a number of warnings were being emitted from using the
deprecated GObject instead of GLib.  Tested on python 3.4 and 3.5.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-08 16:07:35 -06:00
Tony Asleson
3f5629302a lvmdbus: Fix exception during exception handling
Python 3.5 in F24 was throwing the following exception:

Traceback (most recent call last):
  File "/usr/lib/python3.5/site-packages/lvmdbusd/main.py", line 73, in process_request
    req.run_cmd()
  File "/usr/lib/python3.5/site-packages/lvmdbusd/request.py", line 73, in run_cmd
    self.register_error(-1, st)
  File "/usr/lib/python3.5/site-packages/lvmdbusd/request.py", line 123, in register_error
    self._reg_ending(None, error_rc, error)
  File "/usr/lib/python3.5/site-packages/lvmdbusd/request.py", line 115, in _reg_ending
    self.cb_error(self._rc_error)
  File "/usr/lib64/python3.5/site-packages/dbus/service.py", line 669, in <lambda>
    keywords[error_callback] = lambda exception: _method_reply_error(connection, message, exception)
  File "/usr/lib64/python3.5/site-packages/dbus/service.py", line 293, in _method_reply_error
    exception))
  File "/usr/lib64/python3.5/traceback.py", line 136, in format_exception_only
    return list(TracebackException(etype, value, None).format_exception_only())
  File "/usr/lib64/python3.5/traceback.py", line 442, in __init__
    if (exc_value and exc_value.__cause__ is not None
AttributeError: 'str' object has no attribute '__cause__'

This was caused because we were calling the dbus error callback with a
string instead of an actual exception.  On python 3.4 this was apparently
OK, but not with 3.5.  Corrected to pass the exception to error callback.
Change tested on both python 3.4 and 3.5.

Reported-by: Vratislav Podzimek <vpodzime@redhat.com>
Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-08 16:06:45 -06:00
Tony Asleson
6c78175126 lvmdbusd: Set locale
Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-03-04 10:13:04 -06:00
Tony Asleson
081359f6cd lvmdbusd: Do only 1 refresh for Vg.Change()
Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-02-25 16:43:56 -06:00
Tony Asleson
f1bc68beb4 lvmdbusd: Reduce unnecessary state refreshes
When we use udev or have lvm call back into the dbus service when a
change occurs, even if that change originated from the dbus service
we end up refreshing the state of the system twice which is not
needed or wanted.  This change handles this case by removing any
pending refreshes in the worker queue if the state of the system
was just updated.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-02-25 16:43:55 -06:00
David Teigland
664fa6d1be lvmlockd: use new sanlock flag to avoid blocking
This flag allows sanlock requests to avoid all blocking.
2016-02-24 10:10:59 -06:00
Zdenek Kabelac
2e168a52b0 coverity: drop unneeded header 2016-02-23 21:40:12 +01:00
Zdenek Kabelac
05cc1b87a9 gcc: cleanup Wunused-const-variable
Use #define instead, since we do not require actually buffer needs
to exists to eliminated new gcc6 warning:

clvm.h:53:19: warning: ‘CLVMD_SOCKNAME’ defined but not used
[-Wunused-const-variable]
2016-02-23 14:41:24 +01:00
Zdenek Kabelac
dbc71dc05e gcc: cleanup some sign warnings
When comparing unsigned with int, the comparision is made
as 'unsigned' type, so make it rather explicit which type
is being compared.
2016-02-23 12:25:25 +01:00
Tony Asleson
ecc0406886 lvmdbusd: Remove unlimited retries
Change while to for loop to prevent the daemon from getting stuck
when lvm is messed up.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-02-22 16:07:54 -06:00
Tony Asleson
21034644b6 lvmdbusd: Add env variable to use session bus
export LVMDBUSD_SESSION=True to run on the session bus instead
of the system bus so that we can run the unit test without
installing the dbus conf file.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-02-22 16:07:54 -06:00
Tony Asleson
70d0b210e1 lvmdbusd: background.py, fix stdout parse error
It appears that the output of lvconvert --merge can vary some.  The code
was blowing up as it was trying to parse a line of stdout to retrieve the
% complete, but the line did not have the needed format and an execption
was thrown.  The uncaught exception caused the background thread to exit
without updating the job object, which caused the client to hang forever
waiting.  Added a default exception handler to prevent unhandled execptions
causing hangs and removed the parameter skip_first_line as it's no longer
needed.  The code checks to see if the line can be parsed before doing so.

Signed-off-by: Tony Asleson <tasleson@redhat.com>
2016-02-22 16:07:54 -06:00
Marian Csontos
ec81c42179 lvmdbusd: Fix make install for python >= 3.5
python 3.5 renamed .pyo files to .opt-1.pyc.
2016-02-22 18:16:41 +01:00
David Teigland
c716813651 lvmlockd: invalidate name in lockspace struct after remove
After the lockspace has been successfully removed,
invalidate the name field in the lockspace struct.
The struct remains on the list of lockspaces until
the struct can be freed later.  Until the struct is
freed, its name will prevent another new lockspace
from being created with the same name.
2016-02-22 09:36:35 -06:00
David Teigland
161ae36363 test: add make check_lvmlockd_test
Which runs lvmlockd in --test mode, without
any lock manager.

Also make some adjustments to the check_lvmlockd
variations using the lock managds.
2016-02-22 09:36:35 -06:00
Alasdair G Kergon
1a141e8623 lbmdbusd: Support in-tree testing. 2016-02-18 13:26:08 +00:00
Alasdair G Kergon
6af2ce6244 lvmdbusd: Rename module to lvmdbusd 2016-02-18 01:14:56 +00:00
Alasdair G Kergon
0f353e2053 lvmdbus: resync latest changes with original repo 2016-02-18 00:27:57 +00:00
Alasdair G Kergon
5987562cf9 lvmdbus: Add new daemon. 2016-02-17 23:53:35 +00:00
David Teigland
762b0d697f lvmlockd: don't adopt locks from unused lm
When built without dlm or sanlock support, don't
attempt to adopt locks from that lm.
2016-01-28 09:42:45 -06:00
Zdenek Kabelac
fcbef05aae doc: change fsf address
Hmm rpmlint suggest fsf is using a different address these days,
so lets keep it up-to-date
2016-01-21 12:11:37 +01:00
Alasdair G Kergon
509410bbbc clvmd: Initialise udev.
Since commit 2fc126b00d, the library
code requires udev to be initialised for device scanning and
clvmd can fail to find VGs if devices/external_device_info_source
is set to "udev".
2016-01-20 00:58:09 +00:00
Alasdair G Kergon
c812c2dbc7 locking: Add node parameter to query_resource. 2016-01-19 21:42:22 +00:00
David Teigland
06346eab84 lvmlockd: cosemtic improvements to logging
Also pass our name to sanlock so it appears in
the sanlock status output.
2016-01-18 09:35:21 -06:00
David Teigland
95ead96004 lvmlockd: fix lvb validation for conversion
Make the lvb validation rules for convert match
those for unlock (even though it would be very
unlikely or impossible for convert to deal with
zero lvb.)
2016-01-18 09:35:20 -06:00
David Teigland
1752f5c31e lvmlockd: fixes for test mode
Get the test mode working (lvmlockd running with no
lock manager).
2016-01-14 16:01:29 -06:00
David Teigland
1b1f42b490 lvmlockd: fix exit code
libdaemon uses 1 for success
2016-01-13 16:40:02 -06:00
David Teigland
124b490fe6 lvmlockd: update VG lock version earlier
Have commands send lvmlockd the update message
in vg_write instead of vg_commit, so that it's
not done while LVs are suspended.  If the vg_write
is not committed, and the seqno sent to lvmlockd
is not used, then lvmlockd can detect this when
the next update uses the same seqno.
2015-12-15 16:14:49 -06:00
David Teigland
796461a912 vgrename: use process_each_vg
Use process_each_vg() to lock and read the old VG,
and then call the main vgrename code.

When real VG names are used (not a UUID in place of the
old name), the command still pre-locks the new name
(when strcmp wants it locked first), before calling
process_each_vg on the old name.

In the case where the old name is replaced with a UUID,
process_each_vg now translates that UUID into the real
VG name, which it locks and reads.  In this case, we
cannot do pre-locking to maintain lock ordering because
the old name is unknown.  So, in this case the strcmp
based lock ordering is suppressed and the old name is
always locked first.  This opens a remote chance for
lock ordering conflict between racing vgrenames between
two names where one or both commands use the UUID.
2015-12-14 14:26:47 -06:00
David Teigland
dcb26b5f13 lvmlockd: reconnect to lvmetad if it's restarted
If lvmetad is restarted after lvmlockd has connected
to it, then lvmlockd should reconnect.
2015-12-10 10:50:19 -06:00
Alasdair G Kergon
dcd946e95a dmeventd: Don't trust fifo with wrong attrs.
If an existing fifo has the wrong attributes it cannot be trusted
so we must unlink it and recreate it correctly.
(Replaces 2c8d6f5c90: if the other end of
the fifo already got opened while its mode was insecure, delaying the
chmod isn't going to make any difference!)
2015-12-08 01:48:17 +00:00
Alasdair G Kergon
94dab390ef dmeventd: Extend checks on client socket.
Reinstate and extend checks removed by e1b111b02a.

The code has always assumed that only root has access to the directory
containing the fifos and that they are under the complete control of
dmeventd code.  If anything is found not to be as expected, then open()
should certainly not be attempted!
2015-12-08 00:59:39 +00:00
Zdenek Kabelac
68e2ea11a3 mirror: fix condition
Recent patch tested wrong condition for error
2015-12-01 13:59:20 +01:00
Zdenek Kabelac
86e7894ecc cleanup: use dm_get_status_mirror
Use libdm function to parse mirror status report.
2015-12-01 13:03:16 +01:00
David Teigland
931fede81b hash: change name of new lookup function 2015-11-17 11:59:44 -06:00
David Teigland
485d2ca945 lvmetad: different style for hash functions
In lookup, return a count of entries with the
same key rather than the value from a second
entry with the same key.

Using some slightly different names.
2015-11-17 10:27:16 -06:00
David Teigland
d42cedae58 lvmetad: add error explaining duplicate VG names
When a VG name lookup fails because the name matches
multiple VGs, return an explanatory error.
2015-11-16 16:25:49 -06:00
David Teigland
7d1c9e1d5a lvmetad: fix some error logging
For some errors old_meta may not be set, so
check before logging it.
2015-11-16 15:13:26 -06:00
David Teigland
68c386cce7 lvmetad: use lookup_withval in another location
Simply use lookup_withval right away rather than doing a
standard lookup, checking for the wrong mapping, then
repeating with lookup_withval to get the right mapping.
2015-11-16 15:13:22 -06:00
David Teigland
4a984cabc4 lvmetad: remove unneeded FIXME
This case is now handled correctly.
2015-11-16 13:44:57 -06:00
Stéphane Aulery
376892ddf8 clvmd: remove -v option from getopt
-V is documented, not -v.

Signed-off-by: Stéphane Aulery <saulery@free.fr>
2015-11-16 01:15:03 +01:00
David Teigland
d9295410e9 lvmetad: change the new hash to take data len
If the data len is passed into the hash table
and saved there, then the hash table internals
do not need to assume that the data value is
a string at any point.
2015-11-13 16:54:22 -06:00
David Teigland
46193f4a59 lvmetad: handle duplicate VG names
New hash table functions are added that allow for
multiple entries with the same key.  Use of the
vgname_to_vgid hash table is converted to these
new functions since there are multiple entries
in vgname_to_vgid that have the same key (vgname).

When multiple VGs with the same name exist, commands
that reference only a VG name will fail saying the
VG could not be found (that error message could be
improved.)  Any command that works with the select
option can access one of the VGs with -S vg_uuid=X.
vgrename is a special case that allows the first VG
name arg to be replaced by a uuid, which also works.

(The existing hash table implementation is not well
suited for handling this case, but it works ok with
the new extensions.  Changing lvmetad to use its own
custom hash tables may be preferable at some point.)
2015-11-13 14:56:35 -06:00
David Teigland
43777b551d lvmetad: tail chasing to shut up coverity 2015-11-12 09:35:41 -06:00
Zdenek Kabelac
058725c721 dmeventd: fix check for failing open.
Recent change 2c8d6f5c90
actually droped restart when the reason of failing open is missing
device completely - check for ENOENT now as another reason
to start new dmeventd server  (when there is no systemd to maintain it).
2015-11-11 19:54:55 +01:00
David Teigland
cd937efa77 lvmetad: change recent cleanups
with better alternatives.
2015-11-09 12:18:43 -06:00
David Teigland
d2d5191b78 lvmlockd: reverse some unnecessary checking
These are unnecessary or not useful.
2015-11-09 12:18:43 -06:00
David Teigland
6762eec88c lvmlockd: fix the NO_GL_LS condition
indicating when no global lockspace exists.
2015-11-09 11:33:16 -06:00
Zdenek Kabelac
b1215b7f8c lvmlockd: enforce 64bit arithmetic
Coverity suggest to stay on the 'safe' side and widen operators early
so 64bit offset is made from 64bit arithmentic.
2015-11-09 17:04:10 +01:00
Zdenek Kabelac
d6767d753f lvmlockd: add missing error checks
Detect error from function and report them.
2015-11-09 17:04:10 +01:00
Zdenek Kabelac
c2d5cfbdb8 cleanup: use _ for local function 2015-11-09 17:04:10 +01:00
Zdenek Kabelac
3db5ba08b1 cleanup: drop unneded headers
Coverity notices they are not really needed.
2015-11-09 17:04:10 +01:00
Zdenek Kabelac
32762e2a9c libdaemon: fix passing 32bit values for %d
Since %d is now prohibited for its great confusion,
replace it with  FMTd64 and correctly converted int64_t
parameter.
2015-11-09 17:04:10 +01:00
David Teigland
e207ededd6 lvmlockd: shut up warnings 2015-11-09 10:00:04 -06:00
Zdenek Kabelac
0b5a75c9af lvmetad: restore use of FMTd64
daemon_reply_simple() eats just 64bit numbers - so make it explicit.
2015-11-09 14:02:27 +01:00
Zdenek Kabelac
5aae8de776 lvmetad: maintain seqno as int
Keep seqno as 32bit value.
Also use '_' prefix for local _update_metadata.
2015-11-09 12:21:17 +01:00
Zdenek Kabelac
b1dab26be0 cleanup: coverity quiet
While through all codepaths we never 'read' lock_id unless LCKF_CONVERT,
coverity cannot decrypt this.

As since it's usually better to pass in 'well-defined' data structures
preset lock_id to 0.
2015-11-09 10:22:52 +01:00
Zdenek Kabelac
11151121aa cleanup: use fputs for plain strings
Use fputs() when printing plain string,
easier then fprintf which needs to parse it.

Also check fd before close is >= 0 -
it is - but coverity fail to see it, so eliminate
this false-positive warning.
2015-11-09 10:22:52 +01:00
Zdenek Kabelac
c4c5635870 cleanup: drop unneeded assign
prio is always assigned later.
2015-11-09 10:22:52 +01:00
Zdenek Kabelac
99def8f439 lvmetad: check for pointers not NULL
Check for  arg_vgid_lookup and arg_name_lookup not being NULL.
Drop checking arg_vgid and arg_name for NULL since they
are already dereference earlier - thus mostly must be NOT NULL.

(If that would be possible larger rework of this function would be
required).
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
f66fe2c444 lvmetad: require meta_lookup
Coverity: do not call update_pvid_to_vgid() with
meta_lookup == NULL since it is dereferencing it.
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
91bde0f4a1 lvmetad: check for allocation fail
Coverity: missing check for root != NULL as this pointer is
later dereferenced in add_last_node().
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
9df3069083 clvmd: check for pthread_create status
Coverity: likes to see checked function result.
2015-11-09 10:19:18 +01:00
Zdenek Kabelac
2c8d6f5c90 dmeventd: open fifo in one function
Put calls related to fifo opening into a single function.

Fix  Time-Of-Check-Time-Of-Use and use fstat()
and fchmod() on already opened fd instead of
checking first path and then risking to open something
different.
2015-11-09 10:18:53 +01:00
Zdenek Kabelac
e1b111b02a dmeventd: reorder tou 2015-11-09 10:00:33 +01:00
Zdenek Kabelac
8b857bfdc6 libdevmapper-event: fix fifo leak on error path
Coverity: when _init_client() fails, client fifo could have
been already openned and needs to be closed on error path.
2015-11-09 10:00:13 +01:00
David Teigland
4e6377f5ba lvmetad: refactor and document
update_metadata and pv_found update the cached metadata;
these are both reworked to improve the code, organize it
by each possible state and transition, make it much more
clear what's changing, add more error checking and
handling, and add comments.

The state and content of the cache (hash tables) does not
change (apart from some things that didn't work before),
and the communication to/from commands does not change.
The implementation and organization of the code making
the state changes does change significantly.

One detail related to the content of the cache does change:
different hash tables do not reference the same memory any more;
the target values in each hash table are allocated and freed
individually.
2015-11-03 13:18:27 -06:00
Zdenek Kabelac
099466939e thin: dmeventd plugin check number of failures
If plugin's lvm command execution fails too often (>10 times),
there is no point to torture system more then necessary, just log
and drop monitoring in this case.
2015-10-29 12:14:20 +01:00
Zdenek Kabelac
5d76bdcdbd dmeventd: event string parser handles empty field
Improve event string parser to avoid unneeded alloc+free.

Daemon talk function uses '-' to mark NULL/missing field.
So restore the NULL pointer back on parser.

This should have made old tools like 'dmevent_tool' work again.
As now 'uuid' or 'dso' could become NULL and then be
properly used in _want_registered_device() function.

Since lvm2 always fill these parameters, this change should
have no effect on lvm2.
2015-10-27 11:42:40 +01:00
Zdenek Kabelac
b702d67747 dmevent: fix referencing
Plugin increments DSO refcounter in _alloc_thread_status().
2015-10-26 23:31:23 +01:00
Zdenek Kabelac
1a7bea0f0f cleanup: drop debug 2015-10-23 10:54:13 +02:00
Zdenek Kabelac
b5b2a54834 cleanup: fix gcc compile with older pthread
Older pthread library was missing 'trick'
in pthread_cleanup_pop() which lead to
compilation error:

error: label at end of compound statement

Use explicit ';' to fix it.
2015-10-23 10:54:13 +02:00
Zdenek Kabelac
87a39d8bac dmeventd: ensure filter is empty
On some error path we could have left filter set to some value.
Not a big issue - but lets make reporting correct as soon
as we can.
2015-10-23 10:54:13 +02:00
Zdenek Kabelac
cff1c728d8 dmeventd: debug error paths 2015-10-23 10:54:13 +02:00
Zdenek Kabelac
1a2d0a0c72 cleanup: indents 2015-10-22 22:46:10 +02:00
Zdenek Kabelac
8be60e6a65 cleanup: easier to read code
Avoid using #ifdef code and use 'cmd' instead of 'die'.
2015-10-22 22:46:10 +02:00
Zdenek Kabelac
2af696c32f dmeventd: exit on idle
Implementing exit when 'dmeventd' is idle.
Default idle timeout set to 1 hour - after this time period
dmeventd will cleanly exit.

On systems with 'systemd' - service is automatically started with
next contact on dmeventd communication socket/fifo.

On other systems - new dmeventd starts again when lvm2 command detects
its missing and monitoring is needed.
2015-10-22 22:43:03 +02:00
Zdenek Kabelac
4284ba65eb dmeventd: debug signals 2015-10-22 22:41:12 +02:00
Zdenek Kabelac
aeec62ad19 dmeventd: snapshot plugin unmonitor
Send signal to itself to mark plugin as 'finished' as
the watching rule is no longer usable.
2015-10-22 22:41:12 +02:00
Zdenek Kabelac
12aa56d298 dmeventd: handle signal from plugin
Add support to unmonitor device when monitor recognizes there is
nothing to monitor anymore.

TODO: possibly API change with return value could be also used.
2015-10-22 22:40:07 +02:00
Zdenek Kabelac
9156c5d088 dmeventd: rework locking code
Redesign threading code:

- plugin registration runs within its new created thread for
  improved parallel usage.

- wait task is created just once and used during whole plugin lifetime.

- event thread is based over  'events' filter being set - when
  filter is 0, such thread is 'unused'.

- event loop is  simplified.

- timeout thread is never signaling 'processing' thread.

- pending of events filter cnange is properly reported and
  running event thread is signalled when possible.

- helgrind is not reporting problems.
2015-10-22 22:39:24 +02:00
Zdenek Kabelac
466a1c72b7 cleanup: use enums 2015-10-22 22:36:00 +02:00
Zdenek Kabelac
81e9ab3156 dmeventd: code mode _get_device_status
Move _get_device_status() in code.
Use dm_task_no_flush() function when reading status.
(e.g. none blocking for thins pool)
2015-10-22 22:35:25 +02:00
Zdenek Kabelac
15dbd4b56a dmeventd: minimize locking time for get_imeout
Don't hold lock when creating message (allocating memory).
Thread cannot dissapear as it's only the same thread which
may clean it.
2015-10-22 22:34:30 +02:00
Zdenek Kabelac
e2ea2a8147 dmeventd: drop unneded test
Function is never called without device.uuid.
2015-10-22 22:34:29 +02:00
Zdenek Kabelac
941c6354db dmeventd: wake up timer when setting new timeout 2015-10-22 22:34:29 +02:00
Zdenek Kabelac
02eb000f51 dmeventd: use dm_hold_control_dev
Need here to keep  control device opened while there is 'any' dso
plugin loaded - otherwise there would a race closing controlfd
inside lvm2 plugin while some other monitoring thread would
tried to execute another WAITEVENT task.
2015-10-22 22:34:27 +02:00
Zdenek Kabelac
efc76ca33d dmeventd: move dso handling into single code section
Move all DSO related function in front, so they could be easily
referenced from rest of code.

Add proper error paths with logging and error reporting.

Drop mutex locking when releasing DSO - since DSO is always
allocated and released in main 'event' processing thread.
2015-10-22 22:33:19 +02:00
Zdenek Kabelac
590091a4fa dmeventd: using warning level
When dmevend notices problems, but continues to operate normally
change log level to warning.
2015-10-22 22:33:19 +02:00
Zdenek Kabelac
9488cbdd0b dmeventd: no registering of 0 event mask
Whenever user tries to register 0 mask report this as EINVAL.
2015-10-22 22:33:19 +02:00
Zdenek Kabelac
fa9e41d2e3 dmeventd: thin plugin update
Use  dm_make_percent for percentage calculation like lvm2 command.

Use a single call for resize.
2015-10-22 22:33:07 +02:00
Zdenek Kabelac
6b0bc5b2d9 dmeventd: snapshot plugin device removal
Add #ifdef-ed code to have ability to even remove unusable device.
For now purely experimental.
2015-10-22 22:29:53 +02:00
Zdenek Kabelac
7ff5b03e5e dmeventd: snapshot plugin updates
Improve test for invalid snapshot.

Use dm_make_percent() to manipulate with exactly same percentage
as lvm2 command is using.
2015-10-22 22:29:50 +02:00
Zdenek Kabelac
91350f5c6a dmeventd: mirror plugin update
Don't use  --config  - this requires reload of lvm.conf
2015-10-22 22:28:37 +02:00
Zdenek Kabelac
9c5c9e2355 dmeventd: raid plugin reporting
Fix raid logging introduced with last updating commit.
2015-10-22 22:28:37 +02:00
Zdenek Kabelac
cde12cbe9e dmeventd: lvm2 plugin correctly debug
Fix debug message and report exit when really doing it.
Also add missing '_' to static function.
2015-10-22 22:28:37 +02:00
Zdenek Kabelac
ab6d16a8a5 dmeventd: check for malloc return 2015-10-22 22:28:37 +02:00
David Teigland
73e679f33f man: document possible lvmetad -l values
In the man page and the lvmetad help output.
2015-10-21 13:36:57 -05:00
Ondrej Kozina
f67a52677b lvmetad: add optional timeout option 2015-10-20 15:45:14 +02:00
David Teigland
df34fcdafd lvmlockd: use flag to avoid blocking in sanlock_acquire
If a host failed while holding a sanlock lease,
sanlock_acquire will by default block and wait
for the lease to expire before returning.  We
want it to return with an error so we can retry
instead of blocking, which allows us to process
other lock operations.

(Enclose this in an ifdef until the new flag
appears in a sanlock release.)
2015-10-14 14:39:29 -05:00
Zdenek Kabelac
a6d1c8ac65 dmeventd: use matching function
Respect lvm2_log_fn prototype. The idea of 'reusing' print_log with
plain cast is causing very strange crashes with some older 'gcc'  compilers.
So just do it cleanly...
2015-10-14 14:25:27 +02:00
Zdenek Kabelac
3f1a3b7090 dmeventd: fix missing '!'
During recent code changes '!' was badly converted.
2015-10-13 22:10:47 +02:00
Zdenek Kabelac
ccc39be053 dmeventd: compilable without DEBUG CFLAG
Missed compilability without DEBUG.
2015-10-13 20:59:35 +02:00
Zdenek Kabelac
4b586ad3c2 dmeventd: rework raid plugin
Switch to per-thread mem-pool and lock only for lvm2 calls.
Use libdm parser for raid status line.
2015-10-13 16:02:19 +02:00
Zdenek Kabelac
256e432e78 dmeventd: less locking mirror
Rework mirror plugin to use per-thread mem-pool.
This allows to use locking only for lvm2 library calls.
2015-10-13 16:02:18 +02:00
Zdenek Kabelac
51ff7d5ed8 dmeventd: less locking for snapshot and thin
Use lvm lock for lvm library calls.

dm functions in-use are 'thread-safe' since we use
local per-thread mem pool.
2015-10-13 16:01:41 +02:00
Zdenek Kabelac
49e11102c7 dmeventd: add local mempool for raid and mirror
Using local mempools allows to drop locks when such memory is needed.
2015-10-13 15:56:23 +02:00
Zdenek Kabelac
a11cd2ca2d dmeventd: use new macros to init mempool
For thin and snapshot use new macros
to simplify mempool init and destroy
2015-10-13 15:56:03 +02:00
Zdenek Kabelac
f9926e7e6c dmeventd: introduce macro for init
Simplify commonly used pool creation.
2015-10-13 15:56:03 +02:00
Zdenek Kabelac
76ea01dd20 dmeventd: new initialization of plugin threads
Rework thread creation code to better use resources.

New code will not leak 'timeout' registered thread on error path.

Also if the thread already exist - avoid creation of thread
object and it's later destruction.

If the race is noticed during adding new monitoring thread,
such thread is put on cleanup list and -EEXIST is reported.
2015-10-13 15:55:05 +02:00
Zdenek Kabelac
09a8479cb7 cleanup: move system defines before structs 2015-10-13 15:22:58 +02:00
Zdenek Kabelac
0a633750f1 cleanup: avoid using private
Switch private to user.
2015-10-13 15:22:58 +02:00
Zdenek Kabelac
0e2261dbd1 cleanup: remove multilog 2015-10-13 15:22:57 +02:00
Zdenek Kabelac
842a7a17e3 cleanup: always set nsec 2015-10-13 15:22:57 +02:00
Zdenek Kabelac
8b9533f38f dmeventd: support logging on stdout
Add new supported option '-l' log to stdout/stderr.

It has to be paired with '-f' (foreground run).
2015-10-13 15:22:57 +02:00
Zdenek Kabelac
903e9af1b2 cleanup: static with _
Document ? is printing to stderr.
2015-10-13 15:16:56 +02:00
Zdenek Kabelac
e261af52eb dmeventd: handle snapshot overflow
When snapshot reports overflow, handle it in the same way as Invalid.
Until better ideas are implemented.
2015-10-13 15:15:44 +02:00
Zdenek Kabelac
0e27210308 dmeventd: add more debug messages 2015-10-13 15:15:43 +02:00
Zdenek Kabelac
13086c2523 dmeventd: drop changing logging
As we now use 'unified' logging macro system - we no longer need
to protect from change of logging function pointer - it's set
once at the start of dmeventd and not change anymore
(as lvm2 library no longer interferers here).
2015-10-13 15:15:43 +02:00
Zdenek Kabelac
42a9c8b4a6 dmeventd: enable new logging 2015-10-13 15:13:28 +02:00
Zdenek Kabelac
e50583d721 dmeventd: enable new logging 2015-10-13 15:09:35 +02:00
Zdenek Kabelac
c90363b585 dmeventd: use log_ macros in plugins
Convert from syslog to standard log_() macros in plugins.
2015-10-13 15:09:03 +02:00
Zdenek Kabelac
915f0faac1 dmeventd: syslog -> log_xxxxx
Convert logging of dmeventd to libdm logging macros.
2015-10-13 14:43:30 +02:00
Zdenek Kabelac
0641e3a5fd libdmevent: add dm_event_log
Add dm_event_log() logging function.
Controled by dm_event_log_set().
2015-10-13 14:40:46 +02:00
Zdenek Kabelac
11a084cf42 dmeventd: log all commands with CMD: prefix
For easier grep of all processed command - add CMD: prefix to them.

Use  >>>  for entring command processing
and  <<<  when command is processed.
2015-10-13 14:40:46 +02:00
Zdenek Kabelac
d60794c3a3 dmeventd: fix loging report
msg.cmd holds return value from processing after _do_process_request,
so save it before call.
2015-10-13 14:40:46 +02:00
Zdenek Kabelac
7e1c08bb6a dmeventd: correct ret code for fail
As  _do_register_device() calls  register_device() function that returns
1 success & 0 fail -  ret = -ENOMEM for fail case instead of 0.
2015-10-13 14:38:49 +02:00
Zdenek Kabelac
a61f3c5316 clvmd: update help test
Use <n> for <input> value.
2015-09-22 00:08:44 +02:00
David Teigland
b7410c95cf lvmlockd: unlock lv if command fails before lock completes
If lvmlockd acquires an lv lock for a command, but the
command exits before the reply, then the command has
not activated the lv and lvmlockd should unlock it.
This only applies when the lv was not already locked.

(There will always be a chance that the lv lock is held
while the lv is not active, i.e. if the command fails in
the small window between getting the lv lock and before
doing the activation.  In that case, rerunning the
activation command corrects the inconsistency.)

This commit helps by automatically clearing the
inconsistency (lv locked by not activated) in the most
common case when the lv lock operation is slow to
complete and the command is canceled by the user.

This commit also adds and cleans up references to the
client id in a bunch of log messages, which is useful
to follow processing on each independent lock request.
2015-09-16 15:56:54 -05:00
Zdenek Kabelac
fd3d795b93 makefiles: distclean missing files 2015-09-11 21:52:27 +02:00
David Teigland
6bc3d72a65 lvmlockd: prevent vgremove of dlm VG while lockspace is used
This applies the same rule/logic to dlm VGs that has always
existed for sanlock VGs.  Allowing a dlm VG to be removed
while its lockspace was still running on other hosts largely
worked, but there were difficult problems if another VG with
the same name was recreated.  Forcing the VG lockspace to
be stopped, gives both sanlock and dlm VGs the same behavior.
2015-09-11 14:12:13 -05:00
David Teigland
854a559a49 lvmlockd: remove shortcut for lockspace thread cleanup
This shortcut was added for an odd case that I do not
believe is relevant any more.  Having an alternate
path for lockspace thread cleanup is a complication
that could lead to problems.
2015-09-11 13:20:20 -05:00
David Teigland
18dfbbb150 lvmlockd: optimize dlm global lockspace creation
Set a flag indicating that the dlm global lockspace
exists to avoid going through the entire process of
trying to add it before finding it exists.
2015-09-11 13:20:01 -05:00
David Teigland
3670f095c7 lvmlockd: check all variations of lvb values
The dlm will often lose the lvb content, so we need to
check quite a few possibilities for lvb values that
were not being checked before.

Refactoring was required to pass the entire lvb value
back to the core code instead of the single value.

The only functional change should be detecting new
lvb states where metadata is now invalidated where
it wasn't before.
2015-09-10 09:47:26 -05:00
David Teigland
f11d690967 lvmlockd: flag for internal actions
When an action is created by lvmlockd for itself,
there is no client to send the result to.  Add
the NO_CLIENT flag to the action to skip sending
the result to a client.
2015-09-10 09:47:26 -05:00
David Teigland
15ae237d2c lvmlockd: rename ADOPT_CLIENT_ID
to INTERNAL_CLIENT_ID since it will be
used for more than adopting locks.
2015-09-10 09:47:26 -05:00
Alasdair G Kergon
fb12308416 style: Standardise some error paths. 2015-09-05 23:56:30 +01:00
David Teigland
10ccbc5efa lvmlockd: don't stop lockspace for EREMOVED
Undo the part of the recent EREMOVED change which
automatically stopped the lockspace for a remotely
removed VG.  It didn't always work (would not work
when lvb content was rebuilt in the dlm).  This will
be handled better when the lvb content is controlled
more strictly.
2015-09-04 13:41:38 -05:00
David Teigland
c71af0895d lvmlockctl: update command descriptions and add logging
The one line descriptions will match the man page.

Add a syslog entry when enabling/disabling the global lock
to help debugging.
2015-09-04 11:05:13 -05:00
David Teigland
6240a7639d lvmlockd: fixes for starting dlm global lockspace
Remove the optimization/shortcut for starting the dlm global
lockspace when it was already running.

Reenable automatically starting the dlm global lockspace
when a command attempts to use it and it's not yet started.
This had become disabled at some point.
2015-09-03 16:47:54 -05:00
David Teigland
c27015368b lvmlockd: improve log messages for adding dlm global lockspace 2015-09-03 16:47:54 -05:00
Zdenek Kabelac
2b9843c20b dmeventd: reorder mempool allocation
Since lvm2_init() effectively detects memory leaks
allocate mempool after this initialization
(so it's not reported as leak).
2015-09-03 23:34:36 +02:00
Ondrej Kozina
7bbc128c3d lvmpolld: make lvpoll error messages visible
Previously all stderr messages issued by spawned lvpoll command were reported
as INFO only. This made all such messages invisible in syslog or lvmpolld log
while running default configuration.

All lvpoll stderr messages are loged with WARN priority now and lvpoll
command exiting with retcode != 0 is logged with ERROR priority in
syslog and lvmpolld log
2015-09-02 17:24:26 +02:00
David Teigland
8b6226997e lvmlockd: also use vg name in set_vg_info
Include both the VG uuid and name in the lvmetad
set_vg_info message.  This works around an obscure
problem where the VG uuid in lvmlockd is wrong
when one host removes a dlm VG, then creates a new
VG with the same name.  If the dlm lockspace for
the initial VG was never stopped on another host,
that other host will be using the old uuid in its
lvmetad set_vg_info message.  (That can be
corrected with a larger change, but this is an
effective workaround.)
2015-08-28 14:43:58 -05:00
David Teigland
f0b3e05add lvmetad: also accept vg name for set_vg_info
set_vg_info previously accepted only vg uuid,
now accept both vg uuid and vg name.  If the
uuid is provided, it's used just as before,
but if the uuid is not provided, or if it's
not found, then fall back to using the vg
name if that is provided.
2015-08-28 14:36:48 -05:00
David Teigland
3c1924c9c0 lvmlockd: fix starting dlm global lockspace
lvmlockd would fail to recognize that the global lockspace
failed to start if the dlm wasn't running, so future attempts
to start the dlm global lockspace would do nothing, thinking
it was already running.
2015-08-27 16:00:24 -05:00
David Teigland
e4d5d05119 lvmlockd: remove list of inactive lockspaces
This was only used to return two flags indicating specific
reasons for a lock failure so that a more specific error
message could be printed by the command (lockspace had been
stopped, or lockspace had an error starting.)

Remove the list, given its limited usefulness, the fact it
would easily become inaccurate, and the fact it was causing
misleading error messages.  The error conditions it was meant
to help could be reported differently.
2015-08-27 15:23:14 -05:00
David Teigland
58713d34dd lvmlockd: detect when dlm lvb is invalidated
The lvb content can be lost during dlm recovery,
and we need to detect when this happens to revalidate.
2015-08-27 10:27:24 -05:00
David Teigland
231b7df6cc lvmlockd: improve VG removal for lock_type dlm
This makes lvmlockd removal steps for dlm VGs closely match
sanlock VGs.  Because dlm lockspaces are not required to be
stopped on all hosts before vgremove, there is an extra bit
for dlm lockspaces, where a flag is set in the VG lock lvb
indicating that the VG was removed.  If other hosts happen
to use the VG lock they will see this flag and stop their
lockspace.
2015-08-27 10:27:24 -05:00
David Teigland
521136181b lvmlockd: fix to work around dlm lvb bug
Work around a dlm bug that fails to copy lvb
on a NL->EX conversion.
2015-08-27 10:27:24 -05:00
David Teigland
fda19b55b1 lvmlockd: fix dlm EAGAIN checks
libdlm returns EAGAIN in errno
2015-08-27 10:27:24 -05:00
David Teigland
1fae121b22 lvmlockd: fix sending debug info to lvmlockctl
Single messages sent over unix sockets are limited in
size to /proc/sys/net/core/wmem_max, so send the 1MB
debug buffer in smaller chunks to avoid EMSGSIZE.

Also look for EAGAIN and retry sending for a limited
time when the reader is slower than the writer.

Also shift the location of that code so it's the same
as other requests.
2015-08-20 14:07:11 -05:00
Ferenc Wágner
5476ee8655 cmirrord: avoid resync buffer overflow in LOG_SPRINT
Use snprintf() instead of sprintf() to exclude the possibility of
overflowing the resync history buffers.
2015-08-20 19:06:47 +02:00
Ferenc Wágner
3c396cf1e1 cmirrord: avoid debugging buffer overflow in LOG_SPRINT
Use snprintf() instead of sprintf() to exclude the possibility of
overflowing the debugging history buffers.
2015-08-20 19:06:47 +02:00
Ferenc Wágner
1ea1cb6dc9 cmirrord: fix stack smashing
With clusters larger than 3 nodes, the 32-byte debug buffer in
cpg_join_callback() is too small to contain all the node IDs, because
32-bit identifiers are generally rendered in 10 decimal digits.  No fixed
size is good in all cases, but this is conditionally logged debug info,
so we can simply truncate it.  Double the size, nevertheless.
2015-08-20 19:06:47 +02:00
Ferenc Wágner
92a4b5cc3c cmirrord: add --foreground and --help options. 2015-08-20 19:06:47 +02:00
Ferenc Wágner
c0d6056870 cmirrord: move generic setup from daemonize() to init_all()
Apply pidfile creation, removal and signal setup to foreground processes too.
2015-08-20 19:06:47 +02:00
David Teigland
62a87c84ed lvmlockd: ignore cmd close if no locks were taken
When a command closes its connection, don't waste
time looking for locks to purge if it did not make
any lock requests.
2015-08-19 12:16:04 -05:00
David Teigland
ce2e60ab45 lvmlockd: change log_error to log_debug for non error
It's not uncommon for a command like vgchange -an to
deactivate and unlock LVs that were not active, so
don't lock that as an error.
2015-08-19 11:43:39 -05:00
David Teigland
9c5a85ce24 lvmlockctl: fix debug output 2015-08-18 16:49:33 -05:00
Zdenek Kabelac
abb24370e9 cleanup: move var declaration 2015-08-18 16:05:04 +02:00
Zdenek Kabelac
3a3e17d603 cleanup: check pthread result codes 2015-08-18 15:00:08 +02:00
David Teigland
597de9d586 lvmlockd: add LV name to some debug logging
The LV uuid is used as the lock name, and including
the LV name in some log messages makes it easier to
follow log messages.
2015-08-10 15:07:10 -05:00
David Teigland
6bd5bf3cb5 lvmlockd: fix check for other sanlock lockspaces
The check for other sanlock lockspaces was not checking
that the lockspace type was sanlock, so if dlm lockspaces
were visible, they were wrongly included.
2015-08-10 13:16:04 -05:00
David Teigland
fd1782b5fc lvmlockd: handle loss of sanlock lease storage
This adds the infrastructure, code paths, error reporting,
etc. to handle storage errors, or storage loss, under the
sanlock leases in a VG that is being used.  The loss of
storage means sanlock cannot renew its leases, which means
that the host needs to stop using the shared VG before its
leases expire.

This still requires manually shutting down a VG that has
lost lease storage, e.g. unmounting file systems,
deactivating LVs in the VG.  The next step is to
automatically use a command like blkdeactivate to do that.
2015-08-05 10:21:45 -05:00
Peter Rajnoha
c78033233a coverity: return value check in lvmlockd-dlm
daemons/lvmlockd/lvmlockd-dlm.c:647: check_return: Calling "closedir" without checking return value (as is done elsewhere 13 out of 14 times).
2015-08-04 09:49:29 +02:00
Peter Rajnoha
83541123c8 coverity: fix cppcheck warnings
/lib/log/log.c:88: warning[invalidScanfArgType_int]: %llu in format string (no. 2) requires 'unsigned long long *' but the argument type is 'long long *'.
daemons/lvmlockd/lvmlockd-core.c:791: error[uninitstring]: Dangerous usage of 'version' (strncpy doesn't always null-terminate it).
2015-08-04 09:33:55 +02:00
Peter Rajnoha
46e6b2b86e coverity: fix possible resource leak in lvmpolld-core
/daemons/lvmpolld/lvmpolld-core.c:573: leaked_storage: Variable "cmdargv" going out of scope leaks the storage it points to.
2015-08-04 09:25:47 +02:00
David Teigland
d11f8d4228 lvmlockd: automatically remove the dlm global lockspace
The dlm global lockspace is automatically added when the
first dlm VG lockspace is added.  Reverse this by removing
the dlm global lockspace after the last dlm VG lockspace
is removed.  (Remove old non-working code that did this
based on an old command that could explicitly add/remove
the dlm global lockspace.)
2015-08-03 10:23:01 -05:00
David Teigland
78135c24b4 lvmlockd: small fixes and cleanup for lvmlockctl
. clean up the info output for readability
. remove some internal debug output
. fix the daemon quit option
2015-07-30 10:50:22 -05:00
Peter Rajnoha
f6473baffc toolcontext: add switches to create_toolcontext for connections and filters init
Make it possible to decide whether we want to initialize connections and
filters together with toolcontext creation.

Add "filters" and "connections" fields to struct
cmd_context_initialized_parts and set these in cmd_context.initialized
instance accordingly.

(For now, all create_toolcontext calls do initialize connections and
filters, we'll change that in subsequent patch appropriately.)
2015-07-30 13:54:09 +02:00
Peter Rajnoha
6b0c464a34 refactor: toolcontext: add struct cmd_context_initialized_parts
Add struct cmd_context_initialized_parts to wrap up information
about which cmd context pieces are initialized and add variable
of this struct type into struct cmd_context.

Also, move existing "config_initialized" variable that was directly
part of cmd_context into the new cmd_context.initialized wrapper.

We'll be adding more items into the struct cmd_context_initialized_parts
with subsequent patches...
2015-07-30 13:54:05 +02:00
David Teigland
9aabf441bd vgremove: warn when removing sanlock global lock
When the sanlock VG holding the global lock is removed,
print a warning indicating that the global needs to be
enabled in another sanlock VG.
2015-07-29 14:27:32 -05:00
David Teigland
772b54a08b vgcreate: improve checks for existing global lock
This tries harder to avoid creating duplicate global locks in
sanlock VGs by refusing to create a new sanlock VG with a
global lock if other sanlock VGs exist that may have a gl.
2015-07-29 14:27:32 -05:00
Alasdair G Kergon
fa11ddd7df lvmlockd: Drop -lrt now handled by configure. 2015-07-27 14:53:08 +01:00
Alasdair G Kergon
3934ade5a2 gitignore: Update for in-place build. 2015-07-27 13:18:35 +01:00
Alasdair G Kergon
be66243933 clvmd: Fix freeze if client dies holding locks.
Simply running concurrent copies of 'pvscan | true' is enough to make
clvmd freeze: pvscan exits on the EPIPE without first releasing the
global lock.

clvmd notices the client disappear but because the cleanup code that
releases the locks is triggered from within some processing after the
next select() returns, and that processing can 'break' after doing just
one action, it sometimes never releases the locks to other clients.

Move the cleanup code before the select.
Check all fds after select().
Improve some debug messages and warn in the unlikely event that
select() capacity could soon be exceeded.
2015-07-23 23:10:16 +01:00
David Teigland
57534733b7 lvmlockd: improve check for duplicate global locks
When there are duplicate global locks, check if the gl
is still enabled each time a gl or vg lock is acquired
in the lockspace.  Once one of the duplicates is disabled,
then other hosts will recognize that the issue is resolved
without needing to restart the lockspaces.
2015-07-23 10:39:11 -05:00
David Teigland
85b42d7c95 lvmlockd: improve errors when lvm is built without a lock manager
When lvmlockd is compiled without support for one of the
lock managers (sanlock or dlm), and a command tries to use
one of them, explain that in the error message.
2015-07-17 11:16:18 -05:00
Marian Csontos
738ae4a77f lvmpolld: Fix segfault on 32 bit architectures
Explicit conversions are needed to align writes and reads on the stack.
int64_t is popped from stack while int was pushed.
2015-07-10 16:16:57 +02:00
David Teigland
6294509cc6 coverity: cleanup related to lvmlockd
A couple missing mutex unlock on error bugs.
A bunch of buffer size/termination warnings.
2015-07-09 11:29:28 -05:00
Peter Rajnoha
10d27998b3 coverity: missing return value checks 2015-07-09 15:15:15 +02:00
Peter Rajnoha
a9a7c297ae coverity: missing return value for dm_split_lvm_name pass proper DM name instead of NULL value 2015-07-09 13:11:57 +02:00
Peter Rajnoha
023cf21848 coverity: fix possible invalid dereferences
lib/format1/import-export.c:167: var_deref_op: Dereferencing null pointer "vg->lvm1_system_id"
lib/cache/lvmetad.c:1023: var_deref_op: Dereferencing null pointer "this"
daemons/lvmlockd/lvmlockd-core.c:2659: check_after_deref: Null-checking "act" suggests that it may be null, but it has already been dereferenced on all paths leading to the check
/daemons/lvmetad/lvmetad-core.c:1024: check_after_deref: Null-checking "pvmeta" suggests that it may be null, but it has already been dereferenced on all paths leading to the check
2015-07-09 12:07:41 +02:00
David Teigland
903569d533 lvmlockd: remove log_error instances for normal conditions
There are a number of log_error instances that are replaced
by log_debug because they are not errors.
2015-07-08 15:25:14 -05:00
David Teigland
de13abdfdf lvmlockd: fix unreachable code 2015-07-08 11:02:11 -05:00
David Teigland
2566bdfbc3 lvmlockctl: fix uninitialized names
When formatting and printing info from lvmlockd.
Also fix some new line problems.
2015-07-08 10:58:56 -05:00
Peter Rajnoha
3ec4813ba2 coverity: fix missing initialization
... Using uninitialized value "lockd_state" when calling "lockd_vg"
(even though lockd_vg assigns 0 to the lockd_state, but it looks at
previous state of lockd_state just before that so we need to have
that properly initialized!)

libdm/libdm-report.c:2934: uninit_use_in_call: Using uninitialized value "tm". Field "tm.tm_gmtoff" is uninitialized when calling "_get_final_time".

daemons/lvmlockd/lvmlockctl.c:273: uninit_use_in_call: Using uninitialized element of array "r_name" when calling "format_info_r_action". (just added FIXME as this looks unfinished?)
2015-07-08 14:53:30 +02:00
Peter Rajnoha
71f4fbfbde coverity: fix uninitialized values and other reported problems
daemons/lvmlockd/lvmlockd-core.c:5709: error[uninitStructMember]: Uninitialized struct member: ds.....
daemons/lvmlockd/lvmlockd-core.c:799: error[uninitstring]: Dangerous usage of 'version' (strncpy doesn't always null-terminate it)
daemons/lvmlockd/lvmlockd-core.c:646: error[memleakOnRealloc]: Common realloc mistake: 'pollfd' nulled but not freed upon failure
2015-07-08 14:19:51 +02:00
Peter Rajnoha
16e9b32c2f coverity: fix resource leaks
lib/log/log.c:115: leaked_storage: Variable "st" going out of scope leaks the storage it points to
daemons/lvmpolld/lvmpolld-core.c:573: leaked_storage: Variable "cmdargv" going out of scope leaks the storage it points to
daemons/lvmlockd/lvmlockd-core.c:5341: leaked_handle: Handle variable "fd" going out of scope leaks the handle
daemons/lvmlockd/lvmlockctl.c:575: overwrite_var: Overwriting "able_vg_name" in "able_vg_name = strdup(optarg)" leaks the storage that "able_vg_name" points to
daemons/lvmlockd/lvmlockctl.c:571: overwrite_var: Overwriting "able_vg_name" in "able_vg_name = strdup(optarg)" leaks the storage that "able_vg_name" points to
daemons/lvmlockd/lvmlockctl.c:385: leaked_handle: Handle variable "s" going out of scope leaks the handle
2015-07-08 13:56:06 +02:00
Ondrej Kozina
088ee7618d lvmpolld: fix possible memory corruption with mem debug
if lvm2 is built with debug memory options dm_free() is not
mapped directly to std library's free(). This may cause memory corruption
as a line buffer may get reallocated in getline with realloc.

This is a temporary hotfix. Other debug memory failure needs to
be investigated and explained.
2015-07-07 14:49:53 +02:00
David Teigland
d16332be72 lvmetactl: program to interact with lvmetad
This is not installed; it's only a developer utility
at this point.
2015-07-06 15:54:22 -05:00
Alasdair G Kergon
dfe3eb12d0 include: Standardise around new tool.h. 2015-07-06 17:30:18 +01:00
David Teigland
d3605b81f3 configure: enable building lvmlockd without sanlock or dlm 2015-07-06 11:09:58 -05:00
Alasdair G Kergon
810ab095e6 macros: Wrap PRI with FMT.
Create a set of wrappers with embedded % such as
  #define FMTu64 "%" PRIu64
2015-07-06 15:09:17 +01:00
Alasdair G Kergon
36ce97c625 pre-release 2015-07-03 16:34:40 +01:00
David Teigland
fe70b03de2 Add lvmlockd 2015-07-02 15:42:26 -05:00
David Teigland
bf77f71711 lvmetad: add invalidation method
Add the ability to invalidate global or individual VG metadata.
The invalid state is returned to lvm commands along with the metadata.
This allows lvm commands to detect stale metadata from the cache and
reread the latest metadata from disk (in a subsequent patch.)

These changes do not change the protocol or compatibility between
lvm commands and lvmetad.

Global information
------------------

Global information refers to metadata that is not isolated
to a single VG , e.g. the list of vg names, or the list of pvs.

When an external system, e.g. a locking system, detects that global
information has been changed from another host (e.g. a new vg has been
created) it sends lvmetad the message: set_global_info: global_invalid=1.
lvmetad sets the global invalid flag to indicate that its cached data is
stale.

When lvm commands request information from lvmetad, lvmetad returns the
cached information, along with an additional top-level config node called
"global_invalid".  This new info tells the lvm command that the cached
information is stale.

When an lvm command sees global_invalid from lvmated, it knows it should
rescan devices and update lvmetad with the latest information.  When this
is complete, it sends lvmetad the message: set_global_info:
global_invalid=0, and lvmetad clears the global invalid flag.  Further lvm
commands will use the lvmetad cache until it is invalidated again.

The most common commands that cause global invalidation are vgcreate and
vgextend.  These are uncommon compared to commands that report global
information, e.g. vgs.  So, the percentage of lvmetad replies containing
global_invalid should be very small.

VG information
--------------

VG information refers to metadata that is isolated to a single VG,
e.g. an LV or the size of an LV.

When an external system determines that VG information has been changed
from another host (e.g. an lvcreate or lvresize), it sends lvmetad the
message: set_vg_info: uuid=X version=N.  X is the VG uuid, and N is the
latest VG seqno that was written.  lvmetad checks the seqno of its cached
VG, and if the version from the message is newer, it sets an invalid flag
for the cached VG.  The invalid flag, along with the newer seqno are saved
in a new vg_info struct.

When lvm commands request VG metadata from lvmetad, lvmetad includes the
invalid flag along with the VG metadata.  The lvm command checks for this
flag, and rereads the VG from disk if set.  The VG read from disk is sent
to lvmetad.  lvmetad sees that the seqno in the new version matches the
seqno from the last set_vg_info message, and clears the vg invalid flag.
Further lvm commands will use the VG metadata from lvmetad until it is
next invalidated.
2015-06-23 16:36:33 -05:00
David Teigland
fd1376ebef libdaemon: move compare_config to lib
so it can be used elsewhere.
2015-06-17 13:07:52 -05:00
Zdenek Kabelac
ac6b355978 cleanup: drop double const
Second const is unneeded.
Also always create whole array with MAX elements.
2015-06-15 14:46:44 +02:00
Zdenek Kabelac
edbdbddfb6 dmeventd: better debug
Avoid using 'static' variable within threaded environmnent for debuging.
2015-06-15 14:46:44 +02:00
Zdenek Kabelac
3d5f7f90c8 dmeventd: drop stack 2015-06-15 14:46:44 +02:00
Zdenek Kabelac
9da07f1d3e dmeventd: drop pthread_cancel
Drop unused pthread canceling as well as DEBUG printing in signal
handler.
2015-06-15 14:46:44 +02:00
Petr Rockai
4f91ad64c3 lvmetad: Make it possible to clear the outdated PVs list for a VG. 2015-06-10 16:18:57 +02:00
David Teigland
fd29c7f3a1 lvmetad, lvmpolld: remove DL_LIBS from Makefile
and rdynamic.  They are not needed.
2015-06-08 11:48:22 -05:00
Zdenek Kabelac
d1531ab26d cleanup: gcc warn fix, don't hide pvs() 2015-05-27 11:10:43 +02:00
Ondrej Kozina
1aba262edb lvmpolld: terminate error message with a dot and LF 2015-05-27 11:10:43 +02:00
Ondrej Kozina
c069aff21b lvmpolld: zero errno in before strtoul call
Testing errno value without explicitly setting to
zero in before the strtoul call may lead to
unexpected failures.
2015-05-26 16:27:03 +02:00
Alasdair G Kergon
eeb498627c libdm: Add dm_task_get_errno to return ioctl errno.
There are reports of unexplained ioctl failures when using dmeventd.
An explanation might be that the wrong value of errno is being used.

Change libdevmapper to store an errno set by from dm ioctl() directly
and provide it to the caller through a new dm_task_get_errno() function.

[Replaced f9510548667754d9209b232348ccd2d806c0f1d8]
2015-05-26 15:13:49 +01:00
Petr Rockai
55f3369692 lvmetad: Provide entire pvmeta sections for outdated_pvs. 2015-05-20 19:46:14 +02:00
Petr Rockai
1562cd7320 lvmetad: Attach an outdated_pvs list to vg_lookup replies. 2015-05-20 19:46:13 +02:00
Petr Rockai
da1527d65d lvmetad: Clear the vgid_to_outdated_pvs hash on shutdown. 2015-05-20 19:46:13 +02:00