Commits - linux-cluster/cluster

[CMAN] Update to 4.7

Christine Caulfield • 16 years ago

9584b15

lock_dlm: ignore dlm EBUSY on plock unlock

David Teigland • 16 years ago

1728dd2

gfs-kernel: Workaround fix for bz 444912 (gfs_glock_is_locked_by_me problem)

Abhijith Das • 16 years ago

d47add8

Resolves: bz 436383: GFS filesystem size inconsistent

Bob Peterson • 16 years ago

02b5f62

Added busy member to posix_lock struct

Chris Feist • 16 years ago

77ccf10

Some message gets out of place, but there's no need to panic

David Teigland • 16 years ago

98e82d8

lock_dlm: skip busy plocks in unlock

David Teigland • 16 years ago

1c9932d

[DLM] Add tcp_nodelay option to DLM comms

Christine Caulfield • 16 years ago

16b6a74

Fix for bz #419391. gfs_glock_dq was traversing the gl_holders list without

Benjamin Marzinski • 16 years ago

0413a20

Test commit

Chris Feist • 16 years ago

c22322a

Pulled from RHEL4 branch.

Chris Feist • 16 years ago

2a71581

Unblock signals after fork() so heuristics using signals don't hang

Lon Hohberger • 16 years ago

94d9a57

Always the node name from cluster.conf as the node name.

Patrick Caulfield • 16 years ago

13bc2f0

BZ 360681

Ryan O'Hara • 16 years ago

893a9e5

When checking for old messages, compare against the last ACKed message rather

Patrick Caulfield • 16 years ago

0e7fc12

bz 349001

David Teigland • 16 years ago

451d0d0

Revert broken patch & apply correct patch from 338511

Lon Hohberger • 16 years ago

945ae7f

Fix #338511 - dlm trying to connect from wrong IP

Lon Hohberger • 16 years ago

6810e2d

Fix bugzilla #298861 - apps using magmamsg don't connect from correct source IP on multi-homed hosts

Lon Hohberger • 16 years ago

d390bc0

Fix bugzillas #298831, #298861, #354391 - apps using magmamsg don't connect from correct source IP on multi-homed hosts

Lon Hohberger • 16 years ago

6b00d36

Fix #326531 - fence_xvm using dprintf for debugging can cause corruption

Lon Hohberger • 16 years ago

bfefa44

Only update castparam if it's non-NULL.

Patrick Caulfield • 16 years ago

c0ccfa4

Fix 310981

Lon Hohberger • 16 years ago

834c608

Bug 316031: dm-mirror: incorrect order of mirror presuspend ops caus...

Jonathan Brassow • 16 years ago

c62a257

Merge netfs.sh from RHEL4 branch; adds cifs support to netfs.sh

Lon Hohberger • 16 years ago

b324a7b

bz 199433

David Teigland • 16 years ago

47ac5b8

Backport the fence_bullpap agent from the RHEL51 branch. This version fixes a timeout bug.

Ryan McCabe • 16 years ago

3b7e85a

Clear checks out less than or equal to the level we just checked. #310981

Lon Hohberger • 16 years ago

0338a11

Fix #310981 - status check timers wrong

Lon Hohberger • 16 years ago

9367c06

Bug 290821: cmirror write path appears deadlocked after recovery ...

Jonathan Brassow • 16 years ago

afd486b

Don't overwrite castparam if there is no new ast routine to go with it.

Patrick Caulfield • 16 years ago

7b7bfea

Bug 291521: Cluster mirror can become out-of-sync if nominal I/O overl...

Jonathan Brassow • 16 years ago

2b82d2a

add some info about services output

David Teigland • 16 years ago

12d4dd9

fix for bz240115

James Parsons • 16 years ago

eeb80c6

Fix #204084 - SAP RAs for RHEL4 branch

Lon Hohberger • 16 years ago

dbf5e16

Resolve #252241 - add self_fence support to clusterfs

Lon Hohberger • 16 years ago

83a4f2c

Backport the agent from RHEL5.1 to fix bz205638

Ryan McCabe • 16 years ago

1ae0643

Bug 291521: Cluster mirror can become out-of-sync if nominal I/O overla...

Jonathan Brassow • 16 years ago

58d8c17

Add some transition start/end messages so that support can track things

Patrick Caulfield • 16 years ago

5766451

Bug 257881: Flush/recovery collision leads to deadlock after leg ...

Jonathan Brassow • 16 years ago

5843ec7

Fix for bz258161

James Parsons • 16 years ago

98cf3a9

When closing the device, try very hard to release all the locks. In

Patrick Caulfield • 16 years ago

096dfdb

More makefile fixes to allow for build.

Chris Feist • 16 years ago

bb2f7dd

Added Makefile include location to pick up libdlm.h

Chris Feist • 16 years ago

4aaabd2

add tool dir

David Teigland • 16 years ago

1b1deef

Check for a null lvb pointer before trying to memset the lvb to zero.

David Teigland • 16 years ago

c3320a9

Bugzilla 230803:

Wendy Cheng • 16 years ago

935abb8

'service fenced status' always returns failure (1) even if it is running. fix.

Ryan McCabe • 16 years ago

48c57d0

BUG 239856 Processed: failed server election due to suspended mirror...

Jonathan Brassow • 16 years ago

c245995

Just ignore user lock that's not in convert state during conversion

David Teigland • 16 years ago

d3082e8

bz 199433: NULL pointer dereference in cman:process_messages for cmirror

David Teigland • 16 years ago

b7430c9

Fix for bz#238490 (at least)

Patrick Caulfield • 16 years ago

b7cba88

Fix for bz233428

James Parsons • 16 years ago

77b041f

Clean out 'othercon' structures when the connection is closed.

Patrick Caulfield • 16 years ago

00dc2cd

Fix #246669

Lon Hohberger • 16 years ago

f6c3f6e

Committing for Jim to address bz246675

Ryan McCabe • 16 years ago

d84d359

Clean & deallocation 'othercon' structure when closing a connection.

Patrick Caulfield • 16 years ago

408299b

Bug 238629: dm-cmirror: Remote recovery conflict...

Jonathan Brassow • 16 years ago

991b599

Fix typo (in .h file too!) that refers to nodes rather than locks on queues.

Patrick Caulfield • 16 years ago

afabd31

A little more spiel.

Patrick Caulfield • 16 years ago

37235d9

Document query API

Patrick Caulfield • 16 years ago

23176f4

Josef's change to remove some extra jid locking.

Chris Feist • 16 years ago

b4252b7

Fix some typos and cut-n-pasteos

Patrick Caulfield • 16 years ago

5062767

Install man pages

Patrick Caulfield • 16 years ago

5ff3036

Man pages for libdlm!

Patrick Caulfield • 16 years ago

bd6dfda

Honour the mode parameter to dlm_create_lockspace() even if the

Patrick Caulfield • 16 years ago

d1fa637

The previous change resulted in double quotes around the name; change the

David Teigland • 16 years ago

b1b79ec

quote the resource name when printing

David Teigland • 16 years ago

2a51bb6

<Previous check-in>

Jonathan Brassow • 16 years ago

11de7e4

Separate lockdump (compact, one lock per line) from lockdebug (verbose

David Teigland • 16 years ago

be7a90f

Add dlm_tool to RHEL4 branch to join/leave lockspaces and dump lock

David Teigland • 16 years ago

c4b7d43

Print the owner pid of each lock when dumping /proc/cluster/dlm_locks.

David Teigland • 16 years ago

375b912

Add resource agent for named (fix #213524).

Marek 'marx' Grac • 16 years ago

670806a

Fix #236431

Marek 'marx' Grac • 16 years ago

e66a391

Fix bz#245828 "A TCP connection to DLM port blocks DLM operations"

Patrick Caulfield • 16 years ago

d752754

Bugzilla 245776:

Wendy Cheng • 16 years ago

27a85da

Update to support iLO 2 > 1.29

Ryan McCabe • 16 years ago

30d6ff5

Fix: #245171 (same patch as in #231521 for RHEL5)

Marek 'marx' Grac • 16 years ago

ec83753

Bug #243972 - Wrong usage of variables $pid vs $pid_file

Marek 'marx' Grac • 16 years ago

6d22695

Bugzilla 242720

Wendy Cheng • 17 years ago

e3f195e

From jwhiter@redhat.com in bz 240356:

David Teigland • 17 years ago

8ad18af

Bug 240874

Jonathan Brassow • 17 years ago

e0661ff

Fix #240527

Lon Hohberger • 17 years ago

9ae365e

Bug 235040 is fixed in the 4.6 kernel, no need to work around it now.

Jonathan Brassow • 17 years ago

316e785

If misconfigured, HA LVM + mirroring can cause data corruption. We should

Jonathan Brassow • 17 years ago

993d9b0

People seem to think that they have to setup lvm in rgmanager even though they

Jonathan Brassow • 17 years ago

f7be3bd

Bugzilla 239523:

Wendy Cheng • 17 years ago

ccdfc9f

Resolves: bz 238719: GFS fsck is has problems with resource groups

Robert Peterson • 17 years ago

0f3fd66

2007-05-03 Lon Hohberger <lhh at redhat.com>

Lon Hohberger • 17 years ago

4ca3841

Add a check to the only unchecked use of "node" before calling

Patrick Caulfield • 17 years ago

ebd52c8

Fix 212121, part 5

Lon Hohberger • 17 years ago

c039ca3

Add test case for last fix to 212121

Lon Hohberger • 17 years ago

4491098

Bug 238031: cluster mirrors not handling all recovery/write conflicts

Jonathan Brassow • 17 years ago

734ba04

Bug 199433: NULL pointer dereference in cman:process_messages for cmirro...

Jonathan Brassow • 17 years ago

4bb6fbb

Ancillary fix to bug #223907

Lon Hohberger • 17 years ago

336b262

Bug 236580: [HA LVM]: Bringing site back on-line after failure causes pr...

Jonathan Brassow • 17 years ago

965c4c7

Install udev rules file

Patrick Caulfield • 17 years ago

9936b9a

- printing out a message as DMERR when the case was valid

Jonathan Brassow • 17 years ago

1542ea1

Fixed bz193064 (fence_apc_snmp doesn't support linked switches)

Stanko Kupcevic • 17 years ago

0b8aca1

Fix udev rules for DLM devices

Patrick Caulfield • 17 years ago

ed50102

Remove an overzealous BUG statement.

Jonathan Brassow • 17 years ago

d7e4eae

Bug 235686: Kernel BUG at dm_cmirror_server while recovering region

Jonathan Brassow • 17 years ago

aca170e

Bug 234918 Processed: NMI Watchdog detected LOCKUP while running proces...

Jonathan Brassow • 17 years ago

68ca12e

Bug 235252: cmirror synchronization deadlocked waiting for response fro...

Jonathan Brassow • 17 years ago

73ad72d

Make power on work correctly for RIBCL version 2.22 on both iLO2 and iLO:

Ryan McCabe • 17 years ago

c483269

Bug 234539: multiple streams of I/O can cause system to lock up

Jonathan Brassow • 17 years ago

1bc15e0

Resolves: Bugzilla Bug 235033: gfs_fsck: Bad programmer! You forgot

Robert Peterson • 17 years ago

413801e

ack header flags need to be byteswapped or things go horribly wrong

Patrick Caulfield • 17 years ago

950306c

Remove duplicate file.

Patrick Caulfield • 17 years ago

feacffd

Commit missing test case output

Lon Hohberger • 17 years ago

5035d39

Fix 212121, patch 4. Also, add automated start/stop order checks to src/daemons/tests

Lon Hohberger • 17 years ago

63426e5

Bug 233034: cmirror server failure/migration during GFS I/O causes metad...

Jonathan Brassow • 17 years ago

d8493c1

Resolves: Bugzilla Bug 233167: [CitiStreet] Kernel Panic when spawn

Robert Peterson • 17 years ago

49cca6d

Fix rgmanager's start/stop resource ordering for defined types & undefined types; update test run expected outputs

Lon Hohberger • 17 years ago

eab03ac

Fix #220211, pass 2: ensure timings are accurate and provide multi-master conflict resolution

Lon Hohberger • 17 years ago

ef7540e

Fix #232139

Lon Hohberger • 17 years ago

150b9ff

Make sure all gulm servers keep track of the other gulm servers/clients

Chris Feist • 17 years ago

e26234c

Bug 231230: leg failure on cmirrors causes devices to be stuck in SUSPE...

Jonathan Brassow • 17 years ago

074a33b

Bugzilla 220622: forgot to delete some commented out code in previous

Wendy Cheng • 17 years ago

666ced0

Bugzilla 220622: backport GFS2 statfs() implementation to GFS1.

Wendy Cheng • 17 years ago

ba967ae

bug231408

Jonathan Brassow • 17 years ago

7edc581

Fix 213246

Lon Hohberger • 17 years ago

4984237

Fix help message to refer to script as 'fence_scsi_test'.

Ryan O'Hara • 17 years ago

5f96da9

Resolves: 231149

Lon Hohberger • 17 years ago

63c2afb

Fixed bz193065 (fence_apc_snmp usage has too much whitespace)

Stanko Kupcevic • 17 years ago

3840301

Fix fence_apc_snmp build process

Stanko Kupcevic • 17 years ago

e852820

Added fence_scsi_test script.

Chris Feist • 17 years ago

6c453ca

Make sure fence_scsi is installed.

Chris Feist • 17 years ago

fdc61eb

Make sure we install fence_xvm

Chris Feist • 17 years ago

e402c11

- When a cluster log resumes, the server must re-read the log device.

Jonathan Brassow • 17 years ago

f8a1efc

Fix for bz230134 (can't fence port 1:1 with fence_apc)

Stanko Kupcevic • 17 years ago

bc7753d

Stop cman_tool from trying to restart us while we are shutting things down.

Patrick Caulfield • 17 years ago

1b74cf2

add locking around the log list. There was a small window of opportunity

Jonathan Brassow • 17 years ago

1ba7ded

Resolves: bz 229461: gfs_tool fails to report counters

Robert Peterson • 17 years ago

9b4a000

Fix anonuid/anongid parsing in nfsclient.sh

Lon Hohberger • 17 years ago

d35b172

Resolves: 229338

Lon Hohberger • 17 years ago

df746c3

A previous fix that caused cluster mirrors to reattempt recovery to

Jonathan Brassow • 17 years ago

4d57664

- Support for HP ilo2

Ryan McCabe • 17 years ago

47393b0

Resolves: 228823 229254

Lon Hohberger • 17 years ago

c4db8e7

Bug 217895: lost election results from cmirror server cause mirror ...

Jonathan Brassow • 17 years ago

878692d

Resolves: bz 229200: gfs_fsck stuck in infinite loop.

Robert Peterson • 17 years ago

9698619

Bug 228070: DLM assertion when running GFS I/O during cmirror leg failure

Jonathan Brassow • 17 years ago

674e9a9

Add LVM failover agent; by Jon Brassow

Lon Hohberger • 17 years ago

3a421b4

- Document the -S/passwd_script fence params.

Ryan McCabe • 17 years ago

f62a6b7

Final patch for bugzilla 1904745 - there is a deadlock documented in the

Wendy Cheng • 17 years ago

afe6586

Add member_util.sh to installation

Lon Hohberger • 17 years ago

52b8f0d

Add member_util.sh functions

Lon Hohberger • 17 years ago

27da0ae

Make sure utility stuff is installed in the right place

Lon Hohberger • 17 years ago

143c995

Changes to fix the following bugs:

Jonathan Brassow • 17 years ago

89e7e8a

Support "passwd_script" parameter in python fence agents.

Stanko Kupcevic • 17 years ago

7e8688a

fence_apc_snmp ignores "port" parameter

Stanko Kupcevic • 17 years ago

ab69e69

Make service.sh understand lvm RA type

Lon Hohberger • 17 years ago

c4d3c8a

Add the 'addnodeids' feature to ccs_tool so that people can fix cluster.conf

Patrick Caulfield • 17 years ago

9bcb581

Ignore EPIPE error when sending response. This can happen is, for example,

Ryan O'Hara • 17 years ago

673e5f6

Support the "passwd_script" parameter in the C fence agents.

Ryan McCabe • 17 years ago

0f49b6e

Allow single-instance VxFS failover in the file system RA (#214935)

Lon Hohberger • 17 years ago

865413e

Bugzilla 190475: accidentally passed address of a pointer, instead of pointer,

Wendy Cheng • 17 years ago

8ab6b6d

Update the perl fence agents to take the additional command line option -S <path> or stdin param passwd_script=<path>

Ryan McCabe • 17 years ago

9e0b71a

Bugzilla 220622:

Wendy Cheng • 17 years ago

076316b

Don't query rgmanager if the user only wants a node state

Lon Hohberger • 17 years ago

23133a8

Fix #227507 - missing output of service info in clustat -x

Lon Hohberger • 17 years ago

8485193

Change SCSI_WATCHDOG to WATCHDOG. The script and sysconfig file were not in

Ryan O'Hara • 17 years ago

824f843

- Fix for bug #225337

Jonathan Brassow • 17 years ago

b65f549

- Fixed a build issue on ppc64 cause by a function named debug_info,

Chris Feist • 17 years ago

887b2b2

Fix for bz205457

James Parsons • 17 years ago

323badb

fix for bz214410 - thanks josef

James Parsons • 17 years ago

101e907

New apc agent written in python that supports named outlets and outlet groups, minus perl pain. Addresses bz172179 and bz134489. yee haw

James Parsons • 17 years ago

7475ebe

If no password is specified, pass a "-P ''" to the ipmitool to prevent

Ryan O'Hara • 17 years ago

cc2cda8

Add manual override for fenced to RHEL4 branch; patch is a backport from HEAD branch; fixes 223060

Lon Hohberger • 17 years ago

d823313

Add error reporting if msg_open fails; patch from Josef Whiter

Lon Hohberger • 17 years ago

79d17cc

Fix markup for magma_tool man page; fix from Eric S. Raymond

Lon Hohberger • 17 years ago

5bec33f

Remove debug lines in clustat

Lon Hohberger • 17 years ago

c052401

Added support for watchdog "fencing".

Ryan O'Hara • 17 years ago

ac8038e

Log resource rule & instance collisions when not running rg_test.

Lon Hohberger • 17 years ago

35304a1

Record SCSI persistent reservation key in /var/lock/subsys/scsi_reserve.

Ryan O'Hara • 17 years ago

0a46f57

Patch from Fabio Massimo Di Nitto - Fix portability of getuptime function

Lon Hohberger • 17 years ago

61bb79d

Bugzilla 214239 - trimming glock

Wendy Cheng • 17 years ago

77ad404

Use /proc/uptime by default instead of gettimeofday(2) for internal timings to avoid problems when the clock is reset by NTP

Lon Hohberger • 17 years ago

973a1a1

Fix priority filtering in ocf-shellfuncs (make it honor cluster.conf)

Lon Hohberger • 17 years ago

1042efc

Resolves bugzillas: #213533, #216092, #220211, #223002, #223234/#223240

Lon Hohberger • 17 years ago

f9be9b5

Bugzilla 214239 - trimming glock

Wendy Cheng • 17 years ago

e753a8f

Resolves: bz 223493 gfs_fsck runs slower than previous versions

Robert Peterson • 17 years ago

c5edb8a

Always queue kill messages that need ACKs, so we don't block apps

Patrick Caulfield • 17 years ago

c8a424b

Tell SM when the quorum device comes or goes.

Patrick Caulfield • 17 years ago

7147e62

bugzilla 190475 (rename 3-2):

Wendy Cheng • 17 years ago

2d0b999

Fix race that could panic if cman_kill_node() is called when we are

Patrick Caulfield • 17 years ago

4313a3a

Resolves: bz 222748: gfs_grow gets the rgindex out of order.

Robert Peterson • 17 years ago

3866df8

Bugzilla 190475 (rename 3-1)

Wendy Cheng • 17 years ago

4497c86

Apply patch from Simone Gotti; fixes #222744/#222838

Lon Hohberger • 17 years ago

7b9687d

Include fence_scsi in Makefile.

Ryan O'Hara • 17 years ago

0ed83ea

fence_scsi man page

Ryan O'Hara • 17 years ago

398ef74

Resolves: bz190756: gfs knows of directories which it chooses

Robert Peterson • 17 years ago

3e623bc

- Use copytobin target instead of install target to remain compatible with

Chris Feist • 17 years ago

66cbbe1

fix bz 210362: We don't run through the entire gfs_quota sparse file to do a list operation anymore. We get the layout of the gfs_quota file on disk and only read quota information off the data blocks that are actually in use. Also added functionality to GFS_IOCTL_SUPER to provide the metadata of the hidden quota file.

Abhijith Das • 17 years ago

88237ae

Add scsi fence agent to Makefile.

Ryan O'Hara • 17 years ago

27cacb1

fence_scsi Makefile.

Ryan O'Hara • 17 years ago

cc9631a

Now that the kernel is correctly handling sync state change, we can

Jonathan Brassow • 17 years ago

99c6e3b

Remove extra check_quorum call.

Ryan O'Hara • 17 years ago

7bb2f47

Add timeout (default 120 seconds) to 'fence_tool leave'. If the cluster is

Ryan O'Hara • 17 years ago

c2625ca

Add timeout option for fence_tool leave.

Ryan O'Hara • 17 years ago

dec7b19

Fix #210482: Allow enable using failover domain priorities

Lon Hohberger • 17 years ago

bcb79a6

Resolves: 201396

Lon Hohberger • 17 years ago

2343567

Resolves: 201396

Lon Hohberger • 17 years ago

ee609ef

Don't panic if we try to unlock a plock that's already being unlocked.

David Teigland • 17 years ago

39a9c7c

- Fixed a small cman and fence init script bug which would cause error

Chris Feist • 17 years ago

f8728f2

- Fixed a bug which would cause fenced to fail to execute secondary fence

Chris Feist • 17 years ago

3975445

Allow cluster id to ve set in CCS.

Patrick Caulfield • 17 years ago

4cad1f7

Attempt to register the node in the case where it must perform fence_scsi

Ryan O'Hara • 17 years ago

7bbbdc2

Allow 'stop' to release the reservation if and only if there are no other

Ryan O'Hara • 17 years ago

1ad480b

Fix #217147; patch by Navid Sheikhol-Eslami (navid redhat com)

Lon Hohberger • 17 years ago

b1cc41c

Fix #193603 (part 2; return error code if rgmanager start fails), #210455, #212110, #216774, #212634/#218112.

Lon Hohberger • 17 years ago

b72bd45

Record devices that are successfully registered to /var/run/scsi_reserve.

Ryan O'Hara • 17 years ago

d508d08

Rewrite of get_scsi_devices function.

Ryan O'Hara • 17 years ago

1765e21

Fix success/failure reporting when registering devices at startup.

Ryan O'Hara • 17 years ago

1555f82

Replace /var/lock/subsys/${0##*/} with /var/lock/subsys/scsi_reserve.

Ryan O'Hara • 17 years ago

167b28a

Fix incorrect variable name.

Ryan O'Hara • 17 years ago

1003870

Script to help test/configure/report SCSI persistent reservation capabilities.

Ryan O'Hara • 17 years ago

5c2bcdb

Fix copytobin target in xvm directory

Lon Hohberger • 17 years ago

aeadeb6

Add echo statement to force a newline after success/failure.

Ryan O'Hara • 17 years ago

bb911ba

If the drop_count is disabled (set to 0) on a mounted fs, don't

David Teigland • 17 years ago

3e2de3a

- Change the default drop_count from 50,000 to 200,000. The old value

David Teigland • 17 years ago

abe3830

Fix problem when parsing the output from lvs command.

Ryan O'Hara • 17 years ago

7ec871e

Fix variable declaration.

Ryan O'Hara • 17 years ago

214a81c

Fix get_options_stdin routine, which was incorrectly named get_option_stdin.

Ryan O'Hara • 17 years ago

97d083e

Get rid of a number of unnecessary messages, which spray to the console

Jonathan Brassow • 17 years ago

0035ef4

Just a comment update, so I don't forget why I'm checking for

Jonathan Brassow • 17 years ago

91d76ae

Bug 214487: "Attempt to mark a already marked region" messages when ...

Jonathan Brassow • 17 years ago

b4a56ec

Bug 214487: "Attempt to mark a already marked region" messages when ...

Jonathan Brassow • 17 years ago

f14302d

bz 217626

David Teigland • 17 years ago

4f800bb

Remove "self" parameter. This was used to specify the name of the node

Ryan O'Hara • 17 years ago

2c6d46a

Add support for fence_xvm to RHEL4 branch (requires seamonkey-nss-devel and seamonkey-nspr-devel to compile)

Lon Hohberger • 17 years ago

a1e1cd8

Remove LIMIT param from calls to split. Seems that different versions of

Ryan O'Hara • 17 years ago

e6c6ae6

Fence agent for SCSI persistent reservations.

Ryan O'Hara • 17 years ago

4d10bf2

Variable must be quoted in conditional statement.

Ryan O'Hara • 17 years ago

6ab57bc

Support for bz179771

James Parsons • 17 years ago

2a388af

bz179771 support

James Parsons • 17 years ago

46194ce

Update makefile for bz179771

James Parsons • 17 years ago

781ab67

support for new baytech, bz179771

James Parsons • 17 years ago

c467a89

adresses bz179771

James Parsons • 17 years ago

ee7f115

Man page for rsb agent

James Parsons • 17 years ago

0ee1818

Support for RSB fence agent

James Parsons • 17 years ago

b8f7095

Support for rsb agent

James Parsons • 17 years ago

58d6ca8

rsb agent and Makefile

James Parsons • 17 years ago

edae888

Remove debug code.

Ryan O'Hara • 17 years ago

d767482

Bug 217438: scrolling kernel requests to mark mirror regions

Jonathan Brassow • 17 years ago

9656917

Bug 214517: hung cmirror operations due to looping mirror region requests

Jonathan Brassow • 17 years ago

d77dbbb

Initial checkin for scsi_reserve init script (RHEL4).

Ryan O'Hara • 17 years ago

fd3cebf

Fix #192117 - patch ported to RHEL4 from head by Lenny Maiorani

Lon Hohberger • 17 years ago

7a50043

Fix error reporting (#202634)

Lon Hohberger • 17 years ago

a04b19d

add support for DRAC5 (bz# 211918)

Ryan McCabe • 17 years ago

73f4f07

Fix #213218

Lon Hohberger • 17 years ago

ec8fd45

Fix comments to match function names.

Ryan O'Hara • 17 years ago

02c12c9

Bugzilla 216209: GFS has been splitting large writes into smaller atomic

Wendy Cheng • 17 years ago

4bbe68f

Resolves: bz208836 - fatal: invalid metadata block

Robert Peterson • 17 years ago

a18772f

Now that SELinux xattr support is in gfs-kernel, fsck needs to handle them.

Ryan O'Hara • 17 years ago

caef1e5

Include sd_freeze_count in counters output.

Ryan O'Hara • 17 years ago

bc0307c

It was previously possible to confuse gnbd by sending a process a signal while

Benjamin Marzinski • 17 years ago

98599e2

- Rather than just printing out "UNKNOWN", we print out the value

Jonathan Brassow • 17 years ago

c38dbda

This is the fix for Bugzilla Bug 213723: [CS4] system crashes

Robert Peterson • 17 years ago

1176b7c

Added SELinux support.

Ryan O'Hara • 17 years ago

fcde46e

Fix #213246 (Patch by Jeff Layton), #213312

Lon Hohberger • 17 years ago

0046910

- Bug 194131 - cluster mirror copy status can give false (0.00) percent

Jonathan Brassow • 17 years ago

b06ad12

This fixes bz 203804, wherein pid files are left around. Thx jlayton.

James Parsons • 17 years ago

5662567

Correct incorrectly applied patch to fix compatibility issue

Lon Hohberger • 17 years ago

8d7adeb

Fix 202637 - error reporting missing from some agents

Lon Hohberger • 17 years ago

34910b3

Compatibility fix for resource agents between linux-cluster and linux-ha

Lon Hohberger • 17 years ago

f00d6ce

Addresses bz175538 by adding 4/I support and increasing the telnet timeout value, and ALSO adds drac 4p support for firmware 1.40 and later

James Parsons • 17 years ago

6c54fd9

Fix for bugzilla bug 164499: Unable to mount loopback images

Robert Peterson • 17 years ago

213695a

Install Resource Agents

Marek 'marx' Grac • 17 years ago

8845b6f

Utilities for Resource Agents

Marek 'marx' Grac • 17 years ago

c8234ca

Bug #204403. Resource agents for Apache, MySQL, OpenLDAP, Postgres 8, Samba and Tomcat 5.

Marek 'marx' Grac • 17 years ago

303da0b

Fix 208656, pass 1

Lon Hohberger • 17 years ago

6aac78d

Fix #206149

Lon Hohberger • 17 years ago

255d6c1

Fix #208115

Lon Hohberger • 17 years ago

ca70b8f

Fix #207144

Lon Hohberger • 17 years ago

4cc2318

Fix segfault due to missing param

Lon Hohberger • 17 years ago

08ba262

Fix failed->disabled state transitions; #208011

Lon Hohberger • 17 years ago

652e139

Bugzilla 205592:

Wendy Cheng • 17 years ago

44b2faf

Fix bug where -f caused segfault

Lon Hohberger • 17 years ago

64544e8

Fix for bugzilla #207351. The individual gserv processes inherit the atexit

Benjamin Marzinski • 17 years ago

accae39

Apply resource-instance-name.patch

Lon Hohberger • 17 years ago

dbd3683

Before choosing a new global id for a new group, a node discovers

David Teigland • 17 years ago

a6f296e

Addendum to bz 200883. If gfs_fsck can't finish initialization,

Robert Peterson • 17 years ago

1586472

Bugzilla 199984: This kmalloc to vmalloc change seems to cause panic in

Wendy Cheng • 17 years ago

72c2686

This is the fix for bugzilla 200883: gfs_fsck segfaults.

Robert Peterson • 17 years ago

fb8ee5a

bz 205285 fix. Committing josef's noquota mount option patch. Also added noquota to gfs_mount man page

Abhijith Das • 17 years ago

a66b08b

fix for bz 206590. F_UNLCK was returning -ENOENT when it didn't find plocks associated with the given inode. Should return 0 now.

Abhijith Das • 17 years ago

82e7e9b

Add a spinlock around the ls_nodes_gone list. The list is modified

David Teigland • 17 years ago

f358cb8

This fixes 206160 (the RHEL4 version of the INRA bug). Gulm now is locked into

Benjamin Marzinski • 17 years ago

b43c98b

fix for bz 206339. Was passing fl_pid instead of fl_owner causing F_GETLK to return conflicts with a process' own locks

Abhijith Das • 17 years ago

612b66a

process_startdone_barrier_new() can be called when sg->sevent is

David Teigland • 17 years ago

e424966

Remove check for small packets as it seems to cause problems with

Patrick Caulfield • 17 years ago

a0c98a8

- remove speed-bump that was put in for debugging

Jonathan Brassow • 17 years ago

183eedd

Fix 146924 - segfault if cman does not give out a membership list.

Lon Hohberger • 17 years ago

11e745c

2006-09-07 Lon Hohberger <lhh at redhat.com>

Lon Hohberger • 17 years ago

a1e46a6

Support for DRAC ERA

James Parsons • 17 years ago

5fa6382

- fix the bugs I've seen so far - mostly related to the recently added

Jonathan Brassow • 17 years ago

9a133ed

rename 'private' to 'privdata' so that it doesn't upset c++

Patrick Caulfield • 17 years ago

43b8894

Don't create lkids of 0. When the lkid counter for bucket 0 of the

David Teigland • 17 years ago

7b29f6f

make cman_dispatch(, CMAN_DISPATCH_ALL) do what it's supposed to do.

Patrick Caulfield • 17 years ago

5d5b9f2

Fix mixup noted in 200776, but use LOG_INFO (as defined in clulog.h) as default rather than switching to notice so as not to change the actual behavior

Lon Hohberger • 17 years ago

a6f555e

fix bz 201325 - Kernel Oops when passing LKF_CANCEL to dlm_ls_unlock_wait

David Teigland • 17 years ago

f22fc38

Merge from RHEL4U4 branch

Lon Hohberger • 17 years ago

049d614

addresses bz191899

James Parsons • 17 years ago

2f25342

addresses bz193065

James Parsons • 17 years ago

5cc1c79

Removed BULL refs from man page

James Parsons • 17 years ago

6b3de96

fix 200449 - status checks wrong

Lon Hohberger • 17 years ago

d622c9b

Bugzilla 199984: Increasing gt_statfs_slots tunable could significantly

Wendy Cheng • 17 years ago

1d938cc

- further tightening for 199826

Jonathan Brassow • 17 years ago

a811db3

while pulling debugging prints from the last patch, I pulled an

Jonathan Brassow • 17 years ago

a04fb8a

Fix for bug:

Jonathan Brassow • 17 years ago

2314084

Add man pages for qdisk

Lon Hohberger • 17 years ago

5657908

- Fix for:

Jonathan Brassow • 17 years ago

3c92432

Fix for bz 198303. F_GETLK was broken, always used to return zero conflicts for local plocks. Also bogus pid was being returned for local locks. Added a new pid field to gfs posix lock to store and return actual pid

Abhijith Das • 17 years ago

e1ecea4

Allow zero votes.

Patrick Caulfield • 17 years ago

2573adf

dispatch agent function was being called with an extra arg

David Teigland • 17 years ago

4746aa8

- If a cluster mirror was removed, and a new one created fast enough, the

Jonathan Brassow • 17 years ago

aa9de4c

- fix for bug 197263

Jonathan Brassow • 17 years ago

c7f133f

Fix gettid #includes for other build systems; patch from Fabio M. Di Nitto

Lon Hohberger • 17 years ago

f773beb

- I was incorrectly decrementing the sync count when a mirror region

Jonathan Brassow • 17 years ago

46d98fa

fixed gnbd_monitor bug with recovering multiple imports per server.

Benjamin Marzinski • 17 years ago

8a39020

If we get a Master-HELLO and we are not the master for this transition then

Patrick Caulfield • 17 years ago

1525113

Fix includes for build on ia64

Lon Hohberger • 17 years ago

8e7d4ca

gnbd can now multipath exported partitions, instead of just whole devices.

Benjamin Marzinski • 17 years ago

4b153c4

Implements 'label' support for qdisk. Uses /proc/partitions for device

Lon Hohberger • 17 years ago

8164766

Add a log_error() for every abnormal error/exit condition in dlm_lock()

David Teigland • 17 years ago

28a21cd

- fix for bug 195610 (renaming a clustered mirror is broken)

Jonathan Brassow • 17 years ago

027ae75

Fix #194491

Lon Hohberger • 17 years ago

bc44509

This addresses bugzilla bug #156009 - gfs fsck needs a good

Robert Peterson • 17 years ago

24319be

Fix #193859

Lon Hohberger • 17 years ago

6fb7069

Work around #193128, clean up NFS bits

Lon Hohberger • 17 years ago

702c925

Allow support for NULL locks in sm module. Set errno accordingly on gulm/dumb modules (#193128)

Lon Hohberger • 17 years ago

acbdd81

Allow NULL locks to be taken and add convert operation; use this when possible in clu_lock

Lon Hohberger • 17 years ago

5c14b8d

Fix inifinite loop causing hangs in other daemons (#194361)

Lon Hohberger • 17 years ago

aa6e461

- pull some more unnecessary prints (and test cvs commit)

Jonathan Brassow • 17 years ago

b580147

- last typo and a message clean-up

Jonathan Brassow • 17 years ago

b7a1cab

- bah! typo.

Jonathan Brassow • 18 years ago

b108c3a

- it was possible under heavy I/O and machine joining/leaving to run into

Jonathan Brassow • 18 years ago

6fda70a

- error messages that really should be debug messages.

Jonathan Brassow • 18 years ago

5f86b6e

this file is out of date, the copy on the web page is the correct one

David Teigland • 18 years ago

1093fad

- downgrade some messages to INFO

Jonathan Brassow • 18 years ago

b871905

Patch from Fabio Massimo Di Nitto: Fix includes

Lon Hohberger • 18 years ago

efb2c72

Log debug messages at DEBUG level (not INFO)

Lon Hohberger • 18 years ago

0363979

Move gfs_fsync() lock state back to LM_ST_SHARED based on yesterday's phone

Wendy Cheng • 18 years ago

2461511

- Fix for bugs: 193596 193706

Jonathan Brassow • 18 years ago

57239ee

- fix for bug 193905

Jonathan Brassow • 18 years ago

ac19919

RHEL4 defaults nfs export to "sync" so nfsd would invokes nfsd_sync_dir()

Wendy Cheng • 18 years ago

06f7e2b

updated gnbd man pages

Benjamin Marzinski • 18 years ago

4bcc62e

I missed header and local stack variable definition in previous check-in.

Wendy Cheng • 18 years ago

8d38252

Fix #193247

Lon Hohberger • 18 years ago

e21b720

* Fix uninitialized struct parts which broke failover in some cases

Lon Hohberger • 18 years ago

a9610ee

Don't crash if a node has no "votes" property, show a default.

Patrick Caulfield • 18 years ago

96c0a1e

We miss c/mtime update in fsync() via previous check-in. This patch fixes

Wendy Cheng • 18 years ago

c8e4fce

Fix build on 64-bit arches

Lon Hohberger • 18 years ago

1d270d5

Add simple scoring/disk-based quorum daemon

Lon Hohberger • 18 years ago

9ad93a6

Adding missing backslash.

Chris Feist • 18 years ago

2b4dcfa

Changes to get apc_snmp to build without errors.

Chris Feist • 18 years ago

718f095

bz 191222 : removed debug printk

Abhijith Das • 18 years ago

6b3cc61

Added missing sharedir directory from defines.mk.input.

Chris Feist • 18 years ago

2ee9ac1

fixes for fence_apc_snmp

James Parsons • 18 years ago

0288d12

added lines for fence_apc_snmp

James Parsons • 18 years ago

3bb5f58

build fix-ups for makefile

James Parsons • 18 years ago

47b9431

U4 patch set #2 - Workaround for #177358. Fixes for #189218, #185388

Lon Hohberger • 18 years ago

46a9f88

genesis of apc_snmp agent

James Parsons • 18 years ago

c861f84

Fix for 182745

James Parsons • 18 years ago

8f793d6

fence_apc_snmp support

Ryan McCabe • 18 years ago

6159e2f

bz191222 fix. When releasing a glock with GL_NOCACHE flag set, care was not taken to ensure that only one holder for the glock remained. This was corrupting the glock and preventing further access to the glock. FLOCKS use this GL_NOCACHE flag. See bugzilla for more information. This patch needs to go through a test cycle to ensure that it doesn't affect other code

Abhijith Das • 18 years ago

f1b3d50

More intelligent fencing call plus /tmp logging

James Parsons • 18 years ago

f7f8cd2

Pushed sleep timeout out additional 5 seconds due to field reports

James Parsons • 18 years ago

f1cf6ea

Fix memory leak

Lon Hohberger • 18 years ago

f34c9c7

- take account of the fact that the sync status can now change

Jonathan Brassow • 18 years ago

f920d5d

merge with head

Ryan McCabe • 18 years ago

2694581

Found a performance issue in gfs_fsync() implementation where GL_SYNC

Wendy Cheng • 18 years ago

1e85d6f

- Update Alpha Commit

Lon Hohberger • 18 years ago

020c294

Pull the libsysfs specific code out of gnbd.

Benjamin Marzinski • 18 years ago

3102f04

If there are no devices defined within a node's method, that method

David Teigland • 18 years ago

a331dc0

Clear comms sequence number for a node when it leaves the cluster.

Patrick Caulfield • 18 years ago

b475642

fix for bz 190392

Abhijith Das • 18 years ago

6676c41

We need to allocated space for 5 ints, rather than 4 when sending a

Patrick Caulfield • 18 years ago

f77188e

Add comment describing potential recovery problem from bz 180538.

David Teigland • 18 years ago

df1f24a

- Fix for bz #183507. We now retry connections to core from ltpx and lt instead

Chris Feist • 18 years ago

f6c76fa

Add missing calls to libcman

Patrick Caulfield • 18 years ago

c9e6e95

Fixed copy - paste error in usage

James Parsons • 18 years ago

2620c4d

Don't try to print local node stuff is 'us' is NULL.

Patrick Caulfield • 18 years ago

d314ab4

bz 198381: ATIX's patch for nodeid CDPN

Abhijith Das • 18 years ago

801a580

Fix for bugzilla bz 179069: gfs_fsck unable to fix file system.

Robert Peterson • 18 years ago

fe22b73

If quit_threads gets set via an incoming message, don't try to carry on.

Patrick Caulfield • 18 years ago

8032a53

Don't try to delete AUTODELETE barriers in timer context as we can't get

Patrick Caulfield • 18 years ago

b7794b4

printk "lockspace already in use"

David Teigland • 18 years ago

50b2fd7

Don't try to unlock a lock if there's no LKB

Patrick Caulfield • 18 years ago

690d931

return 1 if "status" fails.

Patrick Caulfield • 18 years ago

2cf71cf

fixed. Obeys LSB standards for return values

Abhijith Das • 18 years ago

0b1d768

Flush (release) vfs buffer cache on reads to logical volume

Robert Peterson • 18 years ago

bbd7f39

If it took too long for an uncached gnbd recvd process to stop after the server

Benjamin Marzinski • 18 years ago

171e7d2

Fix for bz 186125: gfs_fsck on GFS 6.1TB filesystem gives error and

Robert Peterson • 18 years ago

369747a

Fixed a bug: Page up wasn't re-reading the buffer when wrapping from

Robert Peterson • 18 years ago

820e164

cman init script fix for bz 159783. dlm module is modprobed immediately after cman module. Previously, dlm module was loaded after cman_tool join.

Abhijith Das • 18 years ago

f68f1cd

add support for UIDs to gnbd RHEL4

Benjamin Marzinski • 18 years ago

cba232e

- ccs_tool can seg fault on upgrade bug 186121

Jonathan Brassow • 18 years ago

642ed54

Make cman_dispatch returns the same as libcman in CVS head.

Patrick Caulfield • 18 years ago

6c29915

Add nodeid to /proc/cluster/status

Patrick Caulfield • 18 years ago

cd04820

Fix for bz 182066. Return code was not being propagated while setting default ACLs causing an EPERM everytime

Abhijith Das • 18 years ago

9817f00

Jump command (J) wasn't working in hex mode except for the first screen

Robert Peterson • 18 years ago

7161806

Add quorum device interface for userspace.

Patrick Caulfield • 18 years ago

961ff5f

When address width changes on the left (due to bigger addresses), adjust the cursor position for selection accordingly

Abhijith Das • 18 years ago

8959a7b

fixed pgup/pgdown transitions within blocks that were going crazy because of some boundary conditions

Abhijith Das • 18 years ago

09b8007

Rewrite of gfs_edit to make it user-friendly and more powerful.

Robert Peterson • 18 years ago

0d168f9

Allow "cman_tool leave force" to stop us in mid-tansition, or before joining

Patrick Caulfield • 18 years ago

2e176bc

copy back stuff from -rSTABLE:

Patrick Caulfield • 18 years ago

62fbe17

Copy back the stuff from -rSTABLE:

Patrick Caulfield • 18 years ago

d9368b0

Only release a lockspace if all users have closed it.

Patrick Caulfield • 18 years ago

feec411

Make buffer size sensible.

Patrick Caulfield • 18 years ago

f2384f6

Fix bz#182233

Patrick Caulfield • 18 years ago

2f415e9

Reverting changes for bz 164499 due to locking concerns.

Robert Peterson • 18 years ago

df0cfec

Bugzilla 182057 - patch 3-3: add dump_stack() into gmalloc - previous check

Wendy Cheng • 18 years ago

3797be0

Bugzilla 182057 - patch 3-3: add dump_stack() into gmalloc so we could know

Wendy Cheng • 18 years ago

93f781a

bugzilla 182057 - patch 3-2:

Wendy Cheng • 18 years ago

3ecf13a

bugzilla 182057 - patch 3-1:

Wendy Cheng • 18 years ago

22ad74c

fix for bz178812. ccsd init script and daemon print errors now

Abhijith Das • 18 years ago

04c60eb

- bump disk log version number to match single machine.

Jonathan Brassow • 18 years ago

d250b52

Fix for Bugzilla Bug 178453 – Slow memory leak in /proc/cluster/dlm_dir

Robert Peterson • 18 years ago

0da5518

Joined work of bugzilla 164331 (Abhijith Das) and 178469 (specsfs):

Wendy Cheng • 18 years ago

61bce87

Reset src/gfs/dir.c - accidentally checked in half done code for

Wendy Cheng • 18 years ago

e72cf93

Bugzilla 178469 (specsfs): Properly relay the error return code back

Wendy Cheng • 18 years ago

3677ffa

Addresses interface change in drac_mc firmware version 1.2

James Parsons • 18 years ago

8d1a28a

- Fixes two of the problems causing bz #169171 dealing with host lookups.

Chris Feist • 18 years ago

32233bc

Support for drac 4/I

James Parsons • 18 years ago

fa09422

This patch is part of fix for bugzilla 178469 where the

Wendy Cheng • 18 years ago

325ab43

Fix for bz168698

James Parsons • 18 years ago

99c8bbb

- printk -> DMxxx

Jonathan Brassow • 18 years ago

d334768

- endianness fixes

Jonathan Brassow • 18 years ago

a18d886

fixed typos

James Parsons • 18 years ago

e2bd4c2

- rtn type for get_region_size is uint32_t

Jonathan Brassow • 18 years ago

ce8f75e

- remove some unneeded prints

Jonathan Brassow • 18 years ago

a35b7be

- changes to table line parsing to go along with what is needed by

Jonathan Brassow • 18 years ago

4f3e457

Set default RSA behavior to reboot, and fixed nit in makefile

James Parsons • 18 years ago

09f09e8

added explanation of new auth type switch

James Parsons • 18 years ago

b4a32c8

documented additional commands in ccs_tool

Abhijith Das • 18 years ago

7b976e4

adding pcaulfield's config file editting commands to RHEL4 with some typo fixes

Abhijith Das • 18 years ago

428b7f1

Agent fixes for #178314

Lon Hohberger • 18 years ago

87951dc

fix #179662

Lon Hohberger • 18 years ago

faf83ca

- Make the init script do a cman_tool leave remove on stop. Restarts

Jonathan Brassow • 18 years ago

b68230b

This fixes a problem from bz #173697. gfs_fsck crashed with certain types of

Benjamin Marzinski • 18 years ago

57449a8

Fix 179063 - some options missing from nfsclient option handler

Lon Hohberger • 18 years ago

9e29c89

Fix 178249 - pass 2

Lon Hohberger • 18 years ago

f2ac4e4

This fixes Bugzilla bz 178367. Memory leak when reading from

Robert Peterson • 18 years ago

251b22f

Added fence_bladecenter back into the install list

James Parsons • 18 years ago

87ea57e

Fix broken build

Lon Hohberger • 18 years ago

03ded84

Fix #178249 - debug messages from gulm.so

Lon Hohberger • 18 years ago

bd37db6

bz127042 fix: kill gnbd_monitor when all uncached gnbds have been removed

Abhijith Das • 18 years ago

9ff11f3

When GFS uses direct-io PR and CW locks are mixed together

David Teigland • 18 years ago

c8c1359

This fixes bugzilla 164499 and allows loopback-mounted filesystems to work

Robert Peterson • 18 years ago

a431311

Fix bz#177163

Patrick Caulfield • 18 years ago

7138210

Fix #166109 - random segfault in clurgmgrd. Fix most of 177467 - clustat hang (does not fix the case where the lock manager never responds to a request)

Lon Hohberger • 18 years ago

6610aa7

Fix 178026 - provide lock holder on request from SM. Fix 178080 - implements work around dlm release 177934 by taking a NULL lock on lockspace acquire.

Lon Hohberger • 18 years ago

921b45e

Add CLK_HOLDER flag: Tells the plugin to return an allocated uint64_t; #178024 - also fixes incorrect API documentation in man page

Lon Hohberger • 18 years ago

b915da2

Fixed support for stdin

James Parsons • 18 years ago

c6e2f44

added support for rsa II

James Parsons • 18 years ago

3bdad5e

added rsa II support

James Parsons • 18 years ago

e124d34

Added fence_rsa

James Parsons • 18 years ago

79c197e

Genesis of the rsa II fence agent man page

James Parsons • 18 years ago

29f3148

Genesis of agent for rsa II interface

James Parsons • 18 years ago

29f44cd

- endian fixes so heterogenious clusters can work WTR CCS

Jonathan Brassow • 18 years ago

7826c7e

- commit Lon's patch

Jonathan Brassow • 18 years ago

2f31bde

fix up for var data type

James Parsons • 18 years ago

e16fdeb

fix for bz176375 regression

James Parsons • 18 years ago

d05fc62

Allow non-root users to create the default lockspace.

Patrick Caulfield • 18 years ago

f0a93a8

Don't send out of a multicast socket.

Patrick Caulfield • 18 years ago

a8c3aa1

Don't pass ioctls to SM if we're not in a cluster.

Patrick Caulfield • 18 years ago

6b8c710

break this _single_ line into three:

David Teigland • 18 years ago

95e7a02

Fix 176343 - __builtin_return_address(x) for x>0 is never guaranteed to work

Lon Hohberger • 18 years ago

a5428a6

fix for bz176165

James Parsons • 18 years ago

2c823de

Fix a race where an attempt to unlock a lock in the completion AST

Patrick Caulfield • 18 years ago

23b20fa

Final fix for bz168698, plus additional usage comment for timeout option.

James Parsons • 18 years ago

f8af08f

seq_start needs to initialise the pointer, because we may not be starting form

Patrick Caulfield • 18 years ago

3fc6981

Don't close existing connection if a double-connect is attempted - just

Patrick Caulfield • 18 years ago

e6da9a1

Remove an assertion. Before sending a lock request we were asserting

David Teigland • 18 years ago

2d4b1e8

- In process_cluster_request() replace most of the assertions with an

David Teigland • 18 years ago

0bca9cb

Bugzilla 160568 - add Asynchronous IO implementation into GFS.

Wendy Cheng • 18 years ago

5d7bc1a

bz169087 - changed flag to BW_WAIT in call to write_buf so that changes to superblock are sync-ed to disk immediately

Abhijith Das • 18 years ago

7cc898b

bz169087 - split fill_super_block() into read_super_block() and fill_super_block(); block_mounters() is called between these two functions

Abhijith Das • 18 years ago

8d163e3

- Fixed a bug would cause luck_gulmd to not call waitpid unless SIGCHLD was

Chris Feist • 18 years ago

20bc728

This gfs change (bugzilla 173913) pairs with kernel (bugzilla 173912)

Wendy Cheng • 18 years ago

c61b688

Fix README...

Lon Hohberger • 18 years ago

c474c48

Fix for bz 142849

Benjamin Marzinski • 18 years ago

4fe6260

Fix #175033 part 2 - read in page-size chunks from /proc/cluster/services

Lon Hohberger • 18 years ago

16ae776

Fix bz#173621

Patrick Caulfield • 18 years ago

bfa8011

Bump SM plugin version

Lon Hohberger • 18 years ago

92f31bf

Fix #175033 - magma-plugins incorrect read behavior for /proc/cluster/services > 4096 bytes

Lon Hohberger • 18 years ago

b169f89

Fix #175229 - remove unneeded references to clurmtabd; it is no longer a necessary piece for NFS failover

Lon Hohberger • 18 years ago

cc0fc25

Sync fs.sh with STABLE branch

Lon Hohberger • 18 years ago

b71e115

Implement 175215: Inherit fsid for nfs exports

Lon Hohberger • 18 years ago

d1e6de5

Fix for bz168698

James Parsons • 18 years ago

8b3ff42

Fix #175114 - rgmanager uses wrong stop-order for unspecified resource agents

Lon Hohberger • 18 years ago

e7e2ca7

Fix #175108 - rgmanager storing extraneous info using VF

Lon Hohberger • 18 years ago

8dad54f

Fix #175106 - lsof -b blocks when using gethostbyname causing slow force-unmount when DNS is broken

Lon Hohberger • 18 years ago

8280ac0

Fix #174819 - clustat crashes if ccsd is not running

Lon Hohberger • 18 years ago

10923e8

Fix #173916 - rgmanager log level change requires restart

Lon Hohberger • 18 years ago

9287673

Fix #173526 - Samba Resource Agent

Lon Hohberger • 18 years ago

8471e2e

Fix #171236 - pass 1 - ia64 alignment warnings

Lon Hohberger • 18 years ago

9b6e6ea

Fix #171153 - pass 1 - clustat withholds information if run on multiple members simultaneously

Lon Hohberger • 18 years ago

826332a

Fix #165447 - ip.sh fails when using VLAN on bonded interface

Lon Hohberger • 18 years ago

2d1bc73

This patch fixes a bug where the quota_enforce setting is ignored. Its

Steven Whitehouse • 18 years ago

f6b60c9

Fix for Bugzilla Bug 155304 – gnbd_monitor doesn't correctly reset after an

Robert Peterson • 18 years ago

609e782

gnbd_serv -n

David Teigland • 18 years ago

ddc181c

Give an interface "max-retries" attempts to get fixed after an error before

Patrick Caulfield • 18 years ago

7040140

Allow cman /proc/cluster/config variables to be set via CCS

Patrick Caulfield • 18 years ago

a7b64e0

Add printk to the assertion failure when a nodeid lookup fails.

David Teigland • 18 years ago

6b287a6

Allow scripts to inherit the name attr of a parent in case the script wants to know it (#172310)

Lon Hohberger • 18 years ago

3fe1059

Fix #162605

Lon Hohberger • 18 years ago

ca8d26d

tidy code style

David Teigland • 18 years ago

344b111

Check in Kevin Anderson's mount sync patch:

Wendy Cheng • 18 years ago

333017a

Removed all files for building rpms as they are not used and out of date.

Chris Feist • 18 years ago

0f45525

The i_sem should be re-locked right after gfs_glock_nq_m(). Otherwise, during error returns, the i_sem count will not be correct.

Wendy Cheng • 18 years ago

710f5b1

Removed extraneous -o options from gcc and specified the soname to match

Chris Feist • 18 years ago

ae0c73b

Add printk("WARNING: dlm_emergency_shutdown\n") to start and

David Teigland • 18 years ago

fc6b6eb

Fixes bz172464; adds WTI RPS10 agent to build

James Parsons • 18 years ago

438d0ee

Fix for 172441

Lon Hohberger • 18 years ago

2ac5495

Fix #172401

Lon Hohberger • 18 years ago

621f675

Fix rest of 172178

Lon Hohberger • 18 years ago

7f974d7

Reorder exclusive glock and i_sem locking sequence to avoid read-write deadlock using direct IO on the same node (bugzilla 171488).

Wendy Cheng • 18 years ago

792a9ae

Fix bz #171211

Patrick Caulfield • 18 years ago

2fea6ff

Fix bugs 172177, 172178

Lon Hohberger • 18 years ago

62fd6b5

Removed extra newline added to test cvs email.

Chris Feist • 18 years ago

3d2b534

Added newline to test email script

Chris Feist • 18 years ago

67c56c0

Flush pages into storage in case of DirectIO falling back to BufferIO.

Wendy Cheng • 18 years ago

9d0c766

Apply patch from Axel Thimm to fix bz172066

Lon Hohberger • 18 years ago

21441ae

- fix some build errors

Jonathan Brassow • 18 years ago

10c1d4e

Ensure rgmanager doesn't block SIGSEGV when debug is not enabled.

Lon Hohberger • 18 years ago

a17034b

- add ability to specify the uuid through the dm table.

Jonathan Brassow • 18 years ago

d100b81

- first set of changes to bring cluster mirroring back in sync

Jonathan Brassow • 18 years ago

bd8bd86

- millennium latency fix for 2.6

Jonathan Brassow • 18 years ago

909e846

Fix #171253 - c

Lon Hohberger • 18 years ago

898a8ba

Fix 150346 - Clustat is riddled with usability problems

Lon Hohberger • 18 years ago

1a9aba9

- Fix 171034 - Add monitoring for local and cluster file systems

Lon Hohberger • 18 years ago

df72f7e

- Add rg_test utility to the default build/install

Lon Hohberger • 18 years ago

def9a19

- Check result code and handle failure's in fsck rgrp read code. (#169340)

Chris Feist • 18 years ago

576a773

Fix for bz #168699 (comment grammer/spelling).

Chris Feist • 18 years ago

736ef14

Incorporate patch from 165447

Lon Hohberger • 18 years ago

2c42ec5

fix for bug 169039. GFS no longer checks permissions before truncating a

Benjamin Marzinski • 18 years ago

5cb7242

Fixed bz167217, and handling of DOWNed interfaces

Stanko Kupcevic • 18 years ago

0960fcc

fix for bz #164324. If it took too long to sync the dependent inodes back to

Benjamin Marzinski • 18 years ago

a0938e4

Proper fix for barriers.

Patrick Caulfield • 18 years ago

4f1c36a

Fixed bz167769: fs.sh doesn't do 10 & 20 OCF_CHECK_LEVEL checking

Stanko Kupcevic • 18 years ago

84b03a4

Really gross hack!!!

Benjamin Marzinski • 18 years ago

57d9fa8

Ensure proper linking

Lon Hohberger • 18 years ago

af35bf3

Fix Joe Orton's comments re: calling ld instead of gcc -shared...

Lon Hohberger • 18 years ago

d93095c

Fix barriers as per bz# 166652

Patrick Caulfield • 18 years ago

2000c36

Install .so libraries with 0755 instead of 0644.

Chris Feist • 18 years ago

7449713

Fix 167216 -- ip.sh script errors

Lon Hohberger • 18 years ago

c4002cc

Fix for bz#166154

Benjamin Marzinski • 18 years ago

a639c7f

Restore the iovec to pristine state after kernel_recvmsg()

Patrick Caulfield • 18 years ago

e7dc0aa

fix for bz #165443. Now, if gfs_setattr_simple() is called by a process that

Benjamin Marzinski • 18 years ago

59bdc96

Add fix for bz #156780

Benjamin Marzinski • 18 years ago

3c3fc35

Don't free the lockinfo block if the LKB still exists.

Patrick Caulfield • 18 years ago

4262351

- change table definition parsing to match what agk will include in LVM2

Jonathan Brassow • 18 years ago

266cfec

- fix problem associating a log to a unique id.

Jonathan Brassow • 18 years ago

4c2c7ff

Changed configure script so when we build we run 'cd xxx && configure'

Chris Feist • 18 years ago

31adfc1

Fixed cmirror-kernel directory name

Chris Feist • 18 years ago

25accb3

Merged in changes from STABLE branch:

Chris Feist • 18 years ago

df66859

Pulled in xdr fixes which didn't get pulled from STABLE.

Chris Feist • 18 years ago

dbce386

Fix off-by-one error in find_node_by_nodeid() that can cause an oops in some

Patrick Caulfield • 18 years ago

9145c06

If all interfaces are down, then exit.

Patrick Caulfield • 18 years ago

28e4520

Add support for Dell PowerEdge 1855 to fence_drac (bz 150563). When using

Adam Manthei • 18 years ago

6621b1d

fix 162501

Lon Hohberger • 18 years ago

eff3ecc

fix 164627

Lon Hohberger • 18 years ago

4007761

- fix cmirror bugs

Jonathan Brassow • 18 years ago

39b5c32

fix for bz 161352

Adam Manthei • 18 years ago

a7db091

Fix 159767

Lon Hohberger • 18 years ago

a0fd392

Fix 163651

Lon Hohberger • 18 years ago

f98261d

o make init.d/fenced exit with WARNING instead of FAILED when using

Adam Manthei • 18 years ago

54ba4fb

Removed all files from cmirror directory (all new stuff should be in cmirror-kernel

Chris Feist • 18 years ago

338f09d

Initial add of cmirror-kernel (rename from cmirror

Chris Feist • 18 years ago

4c78700

- back-port cmirror changes to RHEL4 branch

Jonathan Brassow • 18 years ago

91522c5

Fix 159637

Lon Hohberger • 18 years ago

f1d31df

remove kludge that fixed something I can't find or figure out what it

Michael Conrad Tadpol Tilstra • 18 years ago

62224d1

Merged changes from STABLE branch to fix problem caused by gcc 4 fixes.

Chris Feist • 18 years ago

36ad0ea

Added missing fence_bladecenter to the install Makefile.

Chris Feist • 18 years ago

ce09f15

fix 157327

Lon Hohberger • 18 years ago

bdaa8d1

Apply patch from Eric Kerin to fix #162824, fix #162936

Lon Hohberger • 18 years ago

9308411

Fix type causing verify-all to not work properly

Lon Hohberger • 18 years ago

c814e23

Fix for bz #162709 - some variables weren't getting initialized properly

A. J. Lewis • 18 years ago

69faff8

\Fix 162805\

Lon Hohberger • 18 years ago

f319995

Fix refcounting.

Patrick Caulfield • 18 years ago

1246c22

add fence_drac to default build

Adam Manthei • 18 years ago

d759ec7

Build libcman

Patrick Caulfield • 18 years ago

738c7e6

Increase performance by adding local socket communications

Lon Hohberger • 18 years ago

489c6ca

SM should wait for all recoveries to complete before it processes

David Teigland • 18 years ago

7ddf888

The first mounter shouldn't let others mount until others_may_mount()

David Teigland • 18 years ago

826485c

remove old unused patches

David Teigland • 18 years ago

71ce0a1

Support for Dell Remot Access Card III/XT

Adam Manthei • 18 years ago

34dc274

Don't increment the DLM reference count when connecting to an already

Patrick Caulfield • 18 years ago

01f2b92

return locking errors correctly.

Patrick Caulfield • 18 years ago

26b17a1

More consistency

Patrick Caulfield • 18 years ago

fe24db7

Don't start if the user tries to run the cluster over the loopback device.

Patrick Caulfield • 18 years ago

0835112

Fix for bz #160835 - fsck erroneously reporting bitmap corruption

A. J. Lewis • 18 years ago

d3283dc

Fix minor warning in super.c

A. J. Lewis • 18 years ago

2d05841

Prevent new mounters by changing the lock protocol

A. J. Lewis • 18 years ago

f59dd60

Add a function to write out a new superblock

A. J. Lewis • 18 years ago

8923cb0

Don't use _syscall macro, patch from Adam Conrad

Lon Hohberger • 18 years ago

a9f26a5

Fix for bug #160525

A. J. Lewis • 19 years ago

c19e9cb

Create log_at_* macros

A. J. Lewis • 19 years ago

96c031c

- use umask so that permissions on /etc/cluster/cluster.conf are -rw-r-----

Jonathan Brassow • 19 years ago

b01f009

- fix for bug 157094

Jonathan Brassow • 19 years ago

e3e61d3

Fix bug in ip.sh which would match 10.1.1.1 as being the same as 10.1.1.111

Lon Hohberger • 19 years ago

ee2f697

Change the event id field in sm messages from uint16 to uint32 to

David Teigland • 19 years ago

7da5869

Add patch from Frederik Schueler to remove implicit dependency on rdisc

Lon Hohberger • 19 years ago

8a3d597

Added some diagnostic messages for when clients/slaves cannot connect

Michael Conrad Tadpol Tilstra • 19 years ago

9567808

Don't use SM_RETRY around the kmalloc(ATOMIC) in the barrier callback

David Teigland • 19 years ago

e315fa7

Fix SMP race that could lose messages

Patrick Caulfield • 19 years ago

2d4f763

o get rid of duplicate malloc

A. J. Lewis • 19 years ago

cff8459

Clear connect retry count when closing a connection.

Patrick Caulfield • 19 years ago

15a1e0c

Up max_retries to 5 for bz #139738

Patrick Caulfield • 19 years ago

7e20124

Don't assume child nodes exist just because someone asks for them

Lon Hohberger • 19 years ago

15d9f7d

Ask for node name instead of for the existence of children from ccs

Lon Hohberger • 19 years ago

8d189cd

- fix segfault

Jonathan Brassow • 19 years ago

113627b

Don't ignore CHECK state masters when electing a new master.

Patrick Caulfield • 19 years ago

70f1d21

fixup man page.

Michael Conrad Tadpol Tilstra • 19 years ago

42354d6

o Lack of cast broke fsck - now don't use BH_DATA when assigning to

A. J. Lewis • 19 years ago

22f3349

Fix BZ158133:

Ken Preslan • 19 years ago

018e0e6

Fix for example.conf

Lon Hohberger • 19 years ago

37a670d

Fix arg swap problem when reading from stdin

Lon Hohberger • 19 years ago

e77784a

Fix arg parsing

Lon Hohberger • 19 years ago

df5c9c3

Fix bug #129468

Ken Preslan • 19 years ago

a1918e2

Unbreak build.

Ken Preslan • 19 years ago

28f0f54

fix targetted relocation bug when running with gulm

Lon Hohberger • 19 years ago

ff0f8e4

Bull PAP + Bull IPMI-over-LAN support

Lon Hohberger • 19 years ago

15b590a

Fix API change: we no longer get Logged_out after Fenced

Lon Hohberger • 19 years ago

7cdb3ff

Fix fd leak, change resource-group -> service, fix node ID display

Lon Hohberger • 19 years ago

f8ae924

Fix file descriptor leak in services.c

Lon Hohberger • 19 years ago

e582a0e

Need copy_from_user when reading configuration values from /proc.

Patrick Caulfield • 19 years ago

a5a9757

- s/with and/with an/

Jonathan Brassow • 19 years ago

b5b28bd

Fix 157248

Lon Hohberger • 19 years ago

2077329

Pulled in too much from the head.

Ken Preslan • 19 years ago

ecf2f04

Fix diaper mis-initialization.

Ken Preslan • 19 years ago

18afb64

Fix ACL leak.

Ken Preslan • 19 years ago

3e3cdc7

Arbitrary resource tree patch

Lon Hohberger • 19 years ago

3cdaecd

o Fix for fenced portion of bz #155478

A. J. Lewis • 19 years ago

3956172

Again. My fix of a fix of a fix. Initialized some list variables.

Benjamin Marzinski • 19 years ago

ccc5beb

- get cmirror to compile against rhel4 kernels in /lib/modules/*/build

Jonathan Brassow • 19 years ago

78b58bd

This fixes bz #156635. A variable that I allocated statically in my fix for

Benjamin Marzinski • 19 years ago

6f0e81e

add libgulm.a to the all target. (because i forgot.)

Michael Conrad Tadpol Tilstra • 19 years ago

3ab1767

This fixes journal corruption bug 155597. GFS was able to write new data

Benjamin Marzinski • 19 years ago

a4d4176

remove annoying gnbd message

Benjamin Marzinski • 19 years ago

7f328ad

Fix frim Birger

Lon Hohberger • 19 years ago

6cd123d

Fix timeouts for 32-way bull machines

Lon Hohberger • 19 years ago

fb8acbc

get our own varient of in6addr_loopback.

Michael Conrad Tadpol Tilstra • 19 years ago

960aa21

gcc4-isms

Michael Conrad Tadpol Tilstra • 19 years ago

80beeac

gcc4-isms

Michael Conrad Tadpol Tilstra • 19 years ago

2126ae1

- unbreak the dynamic libs.

Michael Conrad Tadpol Tilstra • 19 years ago

ce99236

o Fix compiler error when using GCC4

A. J. Lewis • 19 years ago

dddf9b2

- new Makefile

Jonathan Brassow • 19 years ago

d7655c1

- cmirror is going into U2 - pull code from head.

Jonathan Brassow • 19 years ago

68ef7c0

Fix GCC4 warnings

Lon Hohberger • 19 years ago

9fb8567

use ar not ld for .a

Michael Conrad Tadpol Tilstra • 19 years ago

e2e585a

Add NBB1600 support + fix IPS800[CE] support

Lon Hohberger • 19 years ago

63a310c

Improve logic that delays and reduces fencing. When fenced is recovering

David Teigland • 19 years ago

9b14411

(copy from head)

David Teigland • 19 years ago

ef8f2e5

lock_dlm now exists, and pool doesn't

David Teigland • 19 years ago

dfd7ed0

Fix bonding driver bug

Lon Hohberger • 19 years ago

fc0e3e5

Fix bug #155318

Ken Preslan • 19 years ago

937db66

Changes to the init scripts. BZ's 153739 and 153741.

Adam Manthei • 19 years ago

5a3babf

Simplify (and fix) the locking around the connection array.

Patrick Caulfield • 19 years ago

dd9df3a

- typos

Jonathan Brassow • 19 years ago

c19f04c

-No cman on upgrade.

Jonathan Brassow • 19 years ago

a0d7f48

Add some schedule()s to SM (thanks Dave) so it doesn't get the chance

Patrick Caulfield • 19 years ago

19fd7a3

Put in an extra check of the DEADNODES flag, after processing messages but

Patrick Caulfield • 19 years ago

aa816cb

Say something if sendmsg fails.

Patrick Caulfield • 19 years ago

78ad4b4

Plug a couple of memory leaks

Patrick Caulfield • 19 years ago

74f18d1

Add version number to the start of all dlm messages to help with

David Teigland • 19 years ago

74b9107

simplify dlm_astd handling by removing the wait_queue; and avoids

David Teigland • 19 years ago

cef4942

Updated the Makefile to use 'ln' to create the links instead

Chris Feist • 19 years ago

a477b01

On Fri, Mar 25, 2005 at 03:22:38PM -0800, Daniel McNeil wrote:

David Teigland • 19 years ago

d102639

Fix timeout bugs

Lon Hohberger • 19 years ago

5e95bf9

Fixed configure script so it displays the correct default for --sbindir.

Chris Feist • 19 years ago

a1b1929

backport from HEAD of:

Michael Conrad Tadpol Tilstra • 19 years ago

848f119

back out extraneous change to groups.c

Lon Hohberger • 19 years ago

9e7b10b

add init.d to install set

Lon Hohberger • 19 years ago

e4875bc

fix warning

Lon Hohberger • 19 years ago

b3a94d8

Use service instead of resourcegroup as root resource to match the UI and user expected behavior / terminology

Lon Hohberger • 19 years ago

58ac49e

Fix various bugzillas (see ChangeLog)

Lon Hohberger • 19 years ago

03f83f1

Fix usage message for -n

Patrick Caulfield • 19 years ago

f83a0fd

Print correct error messages if gethostbyname fails

Patrick Caulfield • 19 years ago

bcb95dd

Set join time on local node

Patrick Caulfield • 19 years ago

85612ba

sleep 5 seconds instead of 1 between fence operations

David Teigland • 19 years ago

000670e

updated makefile, so that the gnbd_serv man page installs correctly

Benjamin Marzinski • 19 years ago

343fc5b

Update the gnbd man pages with Bastian Blanks patches, and added a man page

Benjamin Marzinski • 19 years ago

e4936e3

Now, when a server is started with the -n option, you cannot use fence_gnbd

Benjamin Marzinski • 19 years ago

533af30

Fix for bug 151321. By default, gnbd_import and gnbd_serv now get the nodename

Benjamin Marzinski • 19 years ago

9bb6030

Increase size of gethostbyname_r buffer and improve error if it fails.

Patrick Caulfield • 19 years ago

39efc49

making .so

Michael Conrad Tadpol Tilstra • 19 years ago

0c054e4

Slightly more sensible error returns for some join functions too.

Patrick Caulfield • 19 years ago

35a51ca

Improve error messages.

Patrick Caulfield • 19 years ago

4f7b7b9

Make returned error codes more consistent.

Patrick Caulfield • 19 years ago

d606d6a

Fix 151095

Lon Hohberger • 19 years ago

073bdf0

add resource rule printout to rg_test

Lon Hohberger • 19 years ago

61f837f

Clusterfs.sh / fs.sh fixups

Lon Hohberger • 19 years ago

a8dda7c

Show stdin options with -h

Lon Hohberger • 19 years ago

39228c4

When shutting down a lockspace we can remove lkb's from the ast_queue

David Teigland • 19 years ago

f29850e

When gfs does an lm_cancel() we need to do a dlm cancel if that's

David Teigland • 19 years ago

e775bd9

Fix problems with cancelation. We weren't dealing with waiting locks

David Teigland • 19 years ago

7f41901

Properly cast

Lon Hohberger • 19 years ago

0184a35

Properly cast

Lon Hohberger • 19 years ago

4045706

Change to arch-dependent char instead of uint8_t

Lon Hohberger • 19 years ago

3892486

o Slight modification to fsck man page for new fsck

A. J. Lewis • 19 years ago

71054d1

Clean transitionreason after a state transition has finished.

Patrick Caulfield • 19 years ago

95be6c1

Fix dependancies for join_ccs.o

Patrick Caulfield • 19 years ago

7c4eb98

o continue instead of breaking on errors in scan_inode_list()

A. J. Lewis • 19 years ago

1c63038

Fix 150481, part 2

Lon Hohberger • 19 years ago

bd6c59f

Misc. bugfixes

Lon Hohberger • 19 years ago

9036453

Set close-on-exec on dlm file descriptors.

Patrick Caulfield • 19 years ago

36d471d

Fix #150481

Lon Hohberger • 19 years ago

05171ee

If node isn't found in sm_members report an error and don't oops.

David Teigland • 19 years ago

9544817

Was storing entries under the same name, this prevented withdraw from

Michael Conrad Tadpol Tilstra • 19 years ago

1814bdb

Fix for multiple simultaneous leaves not being handled properly

Lon Hohberger • 19 years ago

2c4c3e8

Ignore any NEWLOCKS or NEWLOCKIDS messages from a previous instance

David Teigland • 19 years ago

a410984

Bastian Blank <bastian@waldi.eu.org>'s manpage munging.

Ken Preslan • 19 years ago

d5b8c2f

o Fix the (d)inode_hash_insert() fxns (fixes bz #149706)

A. J. Lewis • 19 years ago

49552d2

o Update to the latest ondisk.h from the s.r.c. gfs kernel source

A. J. Lewis • 19 years ago

d75dd40

Add -h for clustat / clusvcadm

Lon Hohberger • 19 years ago

c5dc812

o increment link count when dealing with bad '.' & '..' entries

A. J. Lewis • 19 years ago

9c53f24

Tidy printks

Patrick Caulfield • 19 years ago

4e3b4a7

Fixed 146672. While it is still possible to see this bug, the problem that I

Benjamin Marzinski • 19 years ago

8e43df7

If we are a new master, don't try to rejoin an old node.

Patrick Caulfield • 19 years ago

1c9098e

Fix part 2 of #150079

Lon Hohberger • 19 years ago

13d5bb5

Zero out cm_addrs when reading /proc/cluster/services for a member list

Lon Hohberger • 19 years ago

775d2a1

Set the socket priority to INTERACTIVE to ensure

Patrick Caulfield • 19 years ago

9afbb06

implement basic recovery policy handling

Lon Hohberger • 19 years ago

9193baf

Implement status/monitor for nfs clients

Lon Hohberger • 19 years ago

97e43d6

Poll ccsd for config updates.

Lon Hohberger • 19 years ago

637a26a

Back out prio calculation change

Lon Hohberger • 19 years ago

4b601b9

#149522 #150067. Part of #149735. See ChangeLog.

Lon Hohberger • 19 years ago

259561f

Possible fix for bug #146711

Michael Conrad Tadpol Tilstra • 19 years ago

a55f037

Enable build for fence_bullpap and fence_ipmilan

Lon Hohberger • 19 years ago

d08bcac

Add IPMI-over-lan and PAP support for Bull NovaScale machines

Lon Hohberger • 19 years ago

242c372

Misc. fixes; see changelog.

Lon Hohberger • 19 years ago

3093705

o Convert remaining (f)printf's to log_*

A. J. Lewis • 19 years ago

e60d140

Don't send sequence number of zero, it causes trouble.

Patrick Caulfield • 19 years ago

61a1b2b

Get rid of spurious "up" in barrier error path.

Patrick Caulfield • 19 years ago

1f9773a

remove "lkb xxxx exists" message which can flood the console/log.

David Teigland • 19 years ago

46f8d2f

Add a bunch of schedule() calls to potentially-long loops in

David Teigland • 19 years ago

7c6446f

Removed extra argument to fix build error.

Chris Feist • 19 years ago

c43d0d0

Remove unfencing since it needs to be reworked and won't be ready

David Teigland • 19 years ago

b6ccf00

uses the -t option to time out wait operations

Adam Manthei • 19 years ago

5d90d1f

Add -t option to join/leave/wait, specifies the maximum amount of

Patrick Caulfield • 19 years ago

d99de27

Release buffers after use

A. J. Lewis • 19 years ago

6548937

Put some locking round membership_task so we don't try to wake up

Patrick Caulfield • 19 years ago

c116663

When locks on the convert queue are granted, we need to try again to grant

David Teigland • 19 years ago

fe7af65

o adjust the gfs ondisk format number to be the same as the one gfs is

A. J. Lewis • 19 years ago

0a4ce01

Check that a device has been passed in (bz #149261)

A. J. Lewis • 19 years ago

52475fe

Recognize and resolve a second form of conversion deadlock. When

David Teigland • 19 years ago

7f899c2

remove the in_cluster() checking in stop. should no longer be needed

Adam Manthei • 19 years ago

9e2b174

o added "-w" option to cman_tool join (bz #147828)

Adam Manthei • 19 years ago

8d62f6b

I was trying to be too economical with code in cman_tool wait,

Patrick Caulfield • 19 years ago

519a00e

Need to invalidate cached file descriptor if it isn't the right type.

Michael Conrad Tadpol Tilstra • 19 years ago

7a4e228

Yet another attempt to fix 148029

Michael Conrad Tadpol Tilstra • 19 years ago

5a9ea64

If a local client logged into gulm, and the client got the nodelist

Michael Conrad Tadpol Tilstra • 19 years ago

6db2f09

partial fix to cman init script to address bug #147828.

Adam Manthei • 19 years ago

20450e2

Added forgetten line in configure script which caused --sharedir

Chris Feist • 19 years ago

053c2f5

removed a few lines that aren't used for anything

Michael Conrad Tadpol Tilstra • 19 years ago

377b9e2

Display "mantis-friendly" membership state in /proc/cluster/status

Patrick Caulfield • 19 years ago

8332032

"The attached patch makes it possible to ask fence_tool to not wait for

David Teigland • 19 years ago

a1127f6

ignore fence_tool's Q option

David Teigland • 19 years ago

3963109

Deal with failed joinconf more sensibly. (that message may have to go though)

Patrick Caulfield • 19 years ago

dad8c1f

remove some non-critical printk's

David Teigland • 19 years ago

0ca2f80

- commit man page changes from Bastian Blank

Jonathan Brassow • 19 years ago

1c1c557

spell deprecated right.

Michael Conrad Tadpol Tilstra • 19 years ago

f20edb0

Fixed gulm so it depended on lock_harness.symvers, not lock_dlm.symvers.

Chris Feist • 19 years ago

ba0049d

The current manpages are written in plain nroff which is not parsable

David Teigland • 19 years ago

149b03e

We were ignoring blocking callbacks for locks being converted

David Teigland • 19 years ago

67f36d6

Blocking asts were being ignored for all locks being converted which

David Teigland • 19 years ago

7660407

Have the lt actually respond to an error packet if it gets one.

Michael Conrad Tadpol Tilstra • 19 years ago

ba06761

o Count blocks used by inodes and update them if counted doesn't match ondisk

A. J. Lewis • 19 years ago

6342d95

Backport from HEAD of cvs

A. J. Lewis • 19 years ago

7e65e5f

fence_node needs a -O to work as expected.

Michael Conrad Tadpol Tilstra • 19 years ago

b2b7495

That last checkin was a red herring.

Patrick Caulfield • 19 years ago

2ba6e3e

Don't clear the IS_OTHERCON flag when shutting down - it can cause socket leaks.

Patrick Caulfield • 19 years ago

dbfbecc

Remember to tell SM if we get down to one node.

Patrick Caulfield • 19 years ago

94150b9

list fence_tool

David Teigland • 19 years ago

d767996

document wait (-w) option

David Teigland • 19 years ago

d4ed4c0

The current manpages are written in plain nroff which is not parsable

David Teigland • 19 years ago

29c8b31

Add option to fence_tool to wait for the node to complete its join and

David Teigland • 19 years ago

e41f6cd

don't complain when we see fence_tool's -w option

David Teigland • 19 years ago

83c0d16

Display node name in /proc/cluster/status

David Teigland • 19 years ago

466c945

- replace instance of ccs_get with ccs_get_list

Jonathan Brassow • 19 years ago

a000e96

- missing break statement, thanks to Bastian Blank

Jonathan Brassow • 19 years ago

bbca060

document -w and -q options for join

David Teigland • 19 years ago

0c18bee

remove multihome setup from man page

David Teigland • 19 years ago

1614032

- add ccs_tool(8) to SEE ALSO in cluster.conf man page

Jonathan Brassow • 19 years ago

236da92

- add -V and -h to ccs_test

Jonathan Brassow • 19 years ago

bb9cc71

put the info that i thought was in the man pages in there.

Michael Conrad Tadpol Tilstra • 19 years ago

168d60a

Don't start the transition timer if we're doing the shortcut single-node

Patrick Caulfield • 19 years ago

1ff2e37

We add NL locks to a resource to implement gfs's hold_lvb(). When

David Teigland • 19 years ago

13372d7

When a conversion request has been sent to a remote master, that

David Teigland • 19 years ago

d796d0c

don't do a dlm_locks_dump when an assertion fails

David Teigland • 19 years ago

d545213

cman_tool now picks as the local nodename whatever name has been entered

David Teigland • 19 years ago

d376c95

Added incdir to the includes so we don't miss include files that aren't installed.

Chris Feist • 19 years ago

08ea0d3

Added include for so we can find magma libraries if they're in a non standard spot.

Chris Feist • 19 years ago

2924f15

Updated top level Makefile to set DESTDIR so programs put there files in

Chris Feist • 19 years ago

4b2c01d

Should be looking in gulm directory not dlm.

Chris Feist • 19 years ago

c1bd460

Fixed build scripts so they work with the broken modpost.

Chris Feist • 19 years ago

861c38e

Add a -w (wait) option to "cman_tool leave".

Patrick Caulfield • 19 years ago

135d553

Clean up changes from last commit related to bz 143487.

David Teigland • 19 years ago

8c2d8a5

- fence_node has changed to better work with gulm (143487)

Jonathan Brassow • 19 years ago

399cf34

Renamed variables to make it easier to pass arguments to them from .spec file.

Chris Feist • 19 years ago

a97a08c

Fixed bug 147602 - gulm doesn't allow 5 servers

Michael Conrad Tadpol Tilstra • 19 years ago

21956b1

fixed log corruption error. if the latest log entry before a crash was at the

Benjamin Marzinski • 19 years ago

fc59244

Be a bit smarter about when to schedule() when reading lots of data

Patrick Caulfield • 19 years ago

f5721d7

remove unused "bulk lookup" function

David Teigland • 19 years ago

ea967be

David Teigland • 19 years ago

5eef3a4

- exit with an error if an invalid votes value is found

David Teigland • 19 years ago

82c7858

Reverted changes to Makefile done for testing.

Chris Feist • 19 years ago

c8461a2

Updated makefiles so we don't taint the kernel.

Chris Feist • 19 years ago

fdeda88

Fix segfault when not running rgmanager

Lon Hohberger • 19 years ago

3a6173d

remove three log_debug lines that are called so frequently during

David Teigland • 19 years ago

cae24d4

In rcom_send_message, return an error from midcomms_send_message

David Teigland • 19 years ago

c7fed3e

Include rg name in clustat -x output

Lon Hohberger • 19 years ago

a239a80

Added out going queues to the local connect in LTPX. Fixes bug 146670.

Michael Conrad Tadpol Tilstra • 19 years ago

2d2f9bc

Don't install header executable

Patrick Caulfield • 19 years ago

742cc37

- Change log_debug() for normal locking operations to log_debug1() where

David Teigland • 19 years ago

0555d9e

- Assert that the recover_list is empty at the start of dlm_dir_clear.

David Teigland • 19 years ago

69fec13

If recovery was aborted during restbl_rsb_update(), some rsb's could be

David Teigland • 19 years ago

68e6ca8

Check quit_threads in a few more places so we don't get blocked

Patrick Caulfield • 19 years ago

5213e2a

Make "1" the default force value for releasing userspace lockspaces

Patrick Caulfield • 19 years ago

3df1fc0

Add some more fields to the lockdump print out.

Ken Preslan • 19 years ago

871bb1a

When running the "gfs_tool <something> <mountpoint>" commands (e.g.

Ken Preslan • 19 years ago

6e90924

Added slibdir into the toplevel make so we don't touch /usr/lib

Chris Feist • 19 years ago

6072479

Probable fix for BZ#146711.

Ken Preslan • 19 years ago

fbd2e70

Updated copyright code

Chris Feist • 19 years ago

711ef87

undo last change (enforce matching nodename/cluster.conf) so we can

David Teigland • 19 years ago

71b5c08

cman_tool now reports an error if the nodename doesn't match exactly

David Teigland • 19 years ago

879cf20

Clean up lock spaces so we can unload dlm module

Lon Hohberger • 19 years ago

18c6e4d

Port from upstream build fix

Lon Hohberger • 19 years ago

7e4254b

Make heartbeat thread exit whren "quit_threads" gets set. This also mean

Patrick Caulfield • 19 years ago

d75d2ff

pending plocks on gulm can now be interrupted.

Michael Conrad Tadpol Tilstra • 19 years ago

adf0b5e

Use plugindir directive; rename plugins to magma_* to match other system libs with plugins (libipt, pam, etc.)

Lon Hohberger • 19 years ago

f6e4361

magma_tool (replaces cpt).

Lon Hohberger • 19 years ago

2dd0635

cpt->magma_tool. Magma_tool additions. Change configure to accept --slibdir (for static libs) so we can separate static/dynamic library installation (e.g. libs in /lib, static libs in /usr/lib)

Lon Hohberger • 19 years ago

bb37c45

fix for bug 146479, was not properly pushing posix range lock

Michael Conrad Tadpol Tilstra • 19 years ago

1cadebd

If we get a position JOINACK then ignore any negative ones that come afterwards

Patrick Caulfield • 19 years ago

543c845

Don't starve processes that are filling buffers.

Patrick Caulfield • 19 years ago

0e205f0

Fix a bug in a memset().

Ken Preslan • 19 years ago

2d5590d

Don't let stale/invalid data leak up to userspace on a read() from a

Ken Preslan • 19 years ago

f4a6e20

Make GFS quieter in error cases.

Ken Preslan • 19 years ago

7ae8dc3

There are a couple of potential problems this should fix related to

David Teigland • 19 years ago

b32a74b

- ccs(7) not ccs(8)

Jonathan Brassow • 19 years ago

7dcef23

Don't call nodeid2con() if we're shutting down, it might allocate a new con.

Patrick Caulfield • 19 years ago

18ba6b5

go back to persisent dlm_recoverd threads (per lockspace)

David Teigland • 19 years ago

295be55

Put some more validation on integers passed in from the commandline.

Patrick Caulfield • 19 years ago

db9f00e

remove cmirror. it doesn't compile and we don't need it on this branch

David Teigland • 19 years ago

fa6b5c9

ignore any wakeup that arrives after the serviced thread exits

David Teigland • 19 years ago

0207ed7

return an error in process_join_stop instead of asserting

David Teigland • 19 years ago

b3f13d2

Add error indices and errors.txt for explanations

Lon Hohberger • 19 years ago

8d26f6b

Add test.c

Lon Hohberger • 19 years ago

d01cc44

Add consistency tests and supporting code

Lon Hohberger • 19 years ago

0022802

Add make check target; make sure we install man pages

Lon Hohberger • 19 years ago

d356710

Allow specification of resource agent path for testing

Lon Hohberger • 19 years ago

f125b56

Move order of make around so that checks on resources are done before internal consistency checks on rgmanager

Lon Hohberger • 19 years ago

e4a45fd

Add building support for separate memory-watcher. Used by the rg_test program in ../daemons for testing only.

Lon Hohberger • 19 years ago

d6f6b90

Fix up make check target

Lon Hohberger • 19 years ago

b94a6a5

Copy fixes from HEAD

Patrick Caulfield • 19 years ago

88f2559

improve chances that withdraw won't get stuck by:

David Teigland • 19 years ago

bb23393

regex fix for gulm and cman check of /etc/cluster/cluster.conf

Adam Manthei • 19 years ago

7a26500

Change all log_all() to either log_debug() or log_error() since

David Teigland • 19 years ago

fdc24a6

Fix dlm_astd hang in bz 145090. Have dlm_astd skip lkb's in the ast

David Teigland • 19 years ago

a684beb

Add make check, ra-api-1-modified.dtd for validating ra metadata

Lon Hohberger • 19 years ago

cc4c6d7

include time in debug output

David Teigland • 19 years ago

6054783

o fixed broken regex

Adam Manthei • 19 years ago

0472251

Add NFS-tests.

Lon Hohberger • 19 years ago

6025650

Include bonded-ethernet link detection

Lon Hohberger • 19 years ago

413e1ef

start gulm when --use_ccs is used only if <gulm> is defined in /etc/cluster/cluster.conf

Adam Manthei • 19 years ago

2f9e111

don't start cman if <gulm> is defined in /etc/cluster/cluster.conf

Adam Manthei • 19 years ago

6438d11

o cleaned up some man of the man pages

Adam Manthei • 19 years ago

526d62a

Return an error from several ioctls if they are called when we are not part of

Patrick Caulfield • 19 years ago

791298b

improve an error message

David Teigland • 19 years ago

2994921

Forgot to update gio types.

Michael Conrad Tadpol Tilstra • 19 years ago

a523ae3

stop being wavery on where tags go.

Michael Conrad Tadpol Tilstra • 19 years ago

e2a0288

'fence_tool monitor' will print debug logging from fenced

David Teigland • 19 years ago

c1c6032

send all debug logging over local socket so fence_tool can monitor

David Teigland • 19 years ago

38622b5

Man pages

Lon Hohberger • 19 years ago

d8ad1c5

- fix the CCS archive -> xml conversion for the gulm section

Jonathan Brassow • 19 years ago

ef2ab85

amd64 fix

Lon Hohberger • 19 years ago

1f7ca60

- better error reporting when there are multiple concurrent updates

Jonathan Brassow • 19 years ago

bfe86aa

Another remove_joiner() needed - this time if the new node does not respond

Patrick Caulfield • 19 years ago

66c11cf

document new actions wait, status, nodes, services

David Teigland • 19 years ago

b50ba16

Add wait options to cman_tool to help with script synchronisation.

David Teigland • 19 years ago

84cee59

Check for valid LKB in find_lock_by_id() rather than later on when we've

Patrick Caulfield • 19 years ago

d86aa31

Apply Derek Anderson's patch to resolve bug 144514:

Adam Manthei • 19 years ago

d476c9c

Change plugin dir to match other system libs which use plugins. Rename plugins accordingly.

Lon Hohberger • 19 years ago

684651a

Change plugin dir to match other system libs which use plugins. Try plugins in alphabetical order. Ignore plugins which do not have user-readable mode bit set.

Lon Hohberger • 19 years ago

67d9399

return error if ioctl(GETNODE) called before we are a cluster member.

Patrick Caulfield • 19 years ago

d1d354a

Remove some redundant code.

Patrick Caulfield • 19 years ago

7b1828b

remove double \n

David Teigland • 19 years ago

0ab7ff8

log more info to syslog to help people see why nodes are fenced

David Teigland • 19 years ago

93ab19a

mergine back changes.

Michael Conrad Tadpol Tilstra • 19 years ago

86c0c50

- don't ignore the completion callback for a canceled lock

David Teigland • 19 years ago

12988e3

- Leave the lockspace when gfs does a withdraw. For now we abandon

David Teigland • 19 years ago

26caa5e

fix freeing of name string before printing in syslog

David Teigland • 19 years ago

8d6ba70

in startup check add a loop around the cman GETNODE ioctl since it

David Teigland • 19 years ago

5be3bc5

Reverted changes to Makefile so kernel installs are done by hand.

Chris Feist • 19 years ago

dc31bc4

Fix for 44945

Lon Hohberger • 19 years ago

886fbf3

- fix bug with update when absolute path to config file is specified.

Jonathan Brassow • 19 years ago

1f150ca

having init.d/cman load the dlm module makes lon happy

Adam Manthei • 19 years ago

2c383af

o cleaned up init.d script to remove excessive `cman_tool join` and

Adam Manthei • 19 years ago

1cad66d

- add ccs_tool man page

Jonathan Brassow • 19 years ago

c60a474

Changed the way we install modules for 2.6.

Chris Feist • 19 years ago

540e58f

- warn every 30 sec (not every 10) if ccsd can not connect to the

Jonathan Brassow • 19 years ago

9c65a67

- ccs_tool now handles update process

Jonathan Brassow • 19 years ago

c4da58b

If the cluster gets down to 1 node and the last leaver left with

Patrick Caulfield • 19 years ago

3dcacca

If we get nominated as master, remove any joining node we may have.

Patrick Caulfield • 19 years ago

2829de2

Changed the order of the library includes. The old way would use the old

Chris Feist • 19 years ago

759e090

bleh, forgot man page.

Michael Conrad Tadpol Tilstra • 19 years ago

1507acc

use lockserver instead of server. less confusing.

Michael Conrad Tadpol Tilstra • 19 years ago

2ef4a80

rolling changes into rhel4 from head.

Michael Conrad Tadpol Tilstra • 19 years ago

fb4ee24

on shutdown, use gulm_tool to determine node liveliness rather than

Adam Manthei • 19 years ago

6c171b3

- address bug #144806 (ccsd not handeling all clu_connect errors on startup appropriately)

Jonathan Brassow • 19 years ago

4d4d31d

- remove initial ccs_tool perl script

Jonathan Brassow • 19 years ago

018c8ee

pid file is now in /var/run/cluster/ccsd.pid, not /var/run/sistina/ccsd.pid

Adam Manthei • 19 years ago

05c9856

Add Magma to the list of services that are ignore on shutdown.

Adam Manthei • 19 years ago

c9df350

Fix bug that mantis saw; we shouldn't log in with newlines

Adam Manthei • 19 years ago

227992d

move the ordering of the scripts. lock_gulmd needs to be called before clvm2.

Adam Manthei • 19 years ago

71d402a

Copy improved messages code to RHEL4 branch.

Patrick Caulfield • 19 years ago

a1c4119

use die() macro so prog_name prefix is used before error messages

David Teigland • 19 years ago

735fc18

yesterday's signal changes broke "fence_tool leave" which requires we

David Teigland • 19 years ago

f4ca775

s/fence_apc/fence_bladecenter/g

Adam Manthei • 19 years ago

88b3bcf

o makes usage of ccs configurable

Adam Manthei • 19 years ago

5cc0111

fence agent make system updates -- only "supported" agents are in the

Adam Manthei • 19 years ago

b2c7794

print name and ip of node on protocol mismatch error.

Michael Conrad Tadpol Tilstra • 19 years ago

d9f0d0d

Removed -Werror so compile works.

Chris Feist • 19 years ago

219254c

added code to query gulms config from gulm_tool

Michael Conrad Tadpol Tilstra • 19 years ago

406c189

Merging head into RHEL4.

Michael Conrad Tadpol Tilstra • 19 years ago

cc7b298

Clean the queued_messages list at shutdown.

Patrick Caulfield • 19 years ago

a0802ba

use sigsuspend instead of pause as the reliable way to wait for a signal

David Teigland • 19 years ago

252a358

change -s to -n in syslog message

David Teigland • 19 years ago

5a2b8b0

Use ccs_get_list instead of ccs_get to prevent infinite looping

David Teigland • 19 years ago

01ab6d6

syncing change to stable

Benjamin Marzinski • 19 years ago

2662c17

IF YOU USE CCS, YOU WILL WANT TO READ THIS:

Jonathan Brassow • 19 years ago

3fa7727

fence_rib is deprecated by fence_ilo

Adam Manthei • 19 years ago

2b1c726

mention fence_bladecenter in man page for fence_xcat

Adam Manthei • 19 years ago

a8d8eeb

diff -u -p -r1.1 fence_brocade.pl

Adam Manthei • 19 years ago

61f2ffc

Log the name of the node fencing is deferred to or say "prior member"

David Teigland • 19 years ago

f422a48

Copy fixes from head to RHEL4 branch

Patrick Caulfield • 19 years ago

a23451d

- fix for bug #137021 ccs doesn't find most recent cluster.conf

Jonathan Brassow • 19 years ago

e491921

- fix bug #133254 (config file regressing)

Jonathan Brassow • 19 years ago

2cfcefc

pull recent changes from HEAD branch to RHEL4 branch.

Michael Conrad Tadpol Tilstra • 19 years ago

68b7b89

Release rsb's semaphore before queue_ast() since queue_ast() can

David Teigland • 19 years ago

969a234

Get rid of -P option. It should be replace with CLVM labeling later.

Ken Preslan • 19 years ago

60697e6

- fix bug 144179

Jonathan Brassow • 19 years ago

a6c8928

- back port bug fixes from head to RHEL4 branch

Jonathan Brassow • 19 years ago

6e50dbb

linux-cluster / cluster

Commits 1763

linux-cluster / cluster

Source Code

Documentation

Commits 1763