similar to: ocfs2 configuration/performance questions...

Hi all, I'm trying to understand and quantify network traffic for a 32 node high performance OCFS2 installation. Part of what I'm doing is attempting to size the network hardware requirements. To do this I need some help understanding the kinds and amounts of traffic generated during "normal" use. Once populated, the system will have a very high READ to WRITE op ratio. I

Mainline commit d4f7e650e55af6b235871126f747da88600e8040 attempts to delay the dlm_thread from sending the drop ref message if the lockres is being migrated. The problem is that we make the dlm_thread wait for the migration to complete. This causes a deadlock as dlm_thread also participates in the lockres migration process. A better fix for the original oss bugzilla#1012 is in testing.

Hi, After problem described in http://oss.oracle.com/pipermail/ocfs2-users/2010- December/004854.html we've upgraded kernels and ocfs2-tools on every node. The present versions are: kernel 2.6.32-bpo.5-amd64 (from debian lenny-backports) ocfs2-tolls 1.4.4-3 (from debian squeeze) We didn't noticed any problems in logs untill last friday, when the whole ocfs2 cluster crashed. We know

Hi, we have recently started to use ocfs2 on some RHEL 5.5 servers (ocfs2-1.4.7) Some days ago, two servers sharing an ocfs2 filesystem, and with quite virtual services, stalled, in what it seems on ocfs2 issue. This are the lines in their messages files: =====node heraclito (0)======================================== /Dec 4 09:15:06 heraclito kernel: o2net: connection to node parmenides

Please review the list of patches being applied to the ocfs2 1.4 tree. All patches list the mainline commit hash. Thanks Sunil

Hi, all, As we use OCFS2, the network is not good. When the converting request message can?t send to the another node, there will be a node hangs up which will still waiting for the dlm. CAS2/logdir/var/log/syslog.1-6778-Sep 16 20:57:16 CAS2 kernel: [516366.623623] o2net: Connection to node CAS1 (num 1) at 10.172.254.1:7100 has been idle for 30.87 secs, shutting it down.

Hello, I'm writing from the otherside of the world from where my systems are, so details are coming in slow. We have a 6TB OCFS2 volume across 20 or so nodes all running OEL5.4 running ocfs2-1.4.4. The system has worked fairly well for the last 6-8 months. Something has happened over the last few weeks which has driven write performance nearly to a halt. I'm not sure how to proceed, and

All, This is hopefully the final drop of the patches for adding global heartbeat to the o2cb stack. The diff from the previous set is here: http://oss.oracle.com/~smushran/global-hb-diff-2010-10-07 Implemented most of the suggestions provided by Joel and Wengang. The most important one was to activate the feature only at the end, Also, got mostly a clean run with checkpatch.pl. Sunil

Hello all, are there any issues concerning ocfs2 (v 1.2.3) and device-mapper-multipath-0.4.5-12.0.RHEL4? I've build a 4 node cluster with two ocfs2 mounts on /dev/dm-0 and /dev/dm-1. Under heavy load the nodes dies after about two minutes reproducably ;-(( Once I've loaded the ocfs2 modules it seems to me that both modules (device-mapper-multipath related as well as ocfs related) are

backgroud: ocfs2 dlm uses a hash table to store dlm_lock_resource objects. the often used lookup is performed on the hash table. problem: for usages that there are huge number of inodes(thus huge number of dlm_lock_resource objects) in a ocfs2 volume, the lookup performance becomes a problem. the lookup holds spin_lock which could put all others cpus into the state of aquring the spinlock. if

Hello! When is it planned ocfs2 1.6 will be available for RHEL? /Kristian

These patches convert the mle list to a hash. The same patches apply on ocfs2 1.4 too. Currently, we use the same number of hash pages for mles and lockres'. This will be addressed in a future patch that will make both of them configurable. Sunil

o2dlm join and leave messages are more than informational as they are required is debugging locking issues. This patch changes them from KERN_INFO to KERN_NOTICE. Signed-off-by: Sunil Mushran <sunil.mushran at oracle.com> --- fs/ocfs2/dlm/dlmdomain.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index

This patch makes the following needlessly global functions static: - stringify_lockname() - dlm_debug_put() Signed-off-by: Adrian Bunk <bunk at kernel.org> --- fs/ocfs2/dlm/dlmdebug.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) fec83555cf8467e5c932fa52177b8f567eb84d94 diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 5f6d858..1b81dcb 100644 ---

Wether the dlm_lock needs to access lvb or not depends on dlm_lock_resource it belongs to. So a new parameter "struct dlm_lock_resource *res" is added to dlm_new_lock() so that we can know if we need to allocate lvb for the dlm_lock. And we have to make the lockres availale for calling dlm_new_lock(). Signed-off-by: Wengang Wang <wen.gang.wang at oracle.com> ---

in dlmrecovery.c:1121, replace 'migrate' to 'migration' to keep the consistency by comparing to other lines with the similar log info in the same file. Signed-off-by: Jeff Liu <jeff.liu at oracle.com> --- fs/ocfs2/dlm/dlmrecovery.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index

The changes from the last drop are: 1. Patch 11 removes struct dlm_lock_name. 2. Patch 12 is an unrelated bugfix. Actually is related to a bugfix that we are retracting in mainline currently. The patch may need more testing. While I did hit the condition in my testing, Marcos hasn't. I am sending it because it can be queued for 2.6.30. Give us more time to test. 3. Patch 13 will be useful

We should check dlm->dlm_state under dlm->spinlock though maybe in this case it doesn't hurt. Signed-off-by: Wengang Wang <wen.gang.wang at oracle.com> --- fs/ocfs2/dlm/dlmdomain.c | 13 ++++++------- 1 files changed, 6 insertions(+), 7 deletions(-) diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 6b5a492..ab82add 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++

#backporting the 3 patches at http://kernel.us.oracle.com/~smushran/srini/ to 1.2. enlarge hash_table capacity to fasten hash_table lookups. Signed-off-by: Wengang wang <wen.gang.wang at oracle.com> -- diff -up ./svnocfs2-1.2/fs/ocfs2/dlm/dlmdebug.c.orig ./svnocfs2-1.2/fs/ocfs2/dlm/dlmdebug.c --- ./svnocfs2-1.2/fs/ocfs2/dlm/dlmdebug.c.orig 2009-04-22 11:00:37.000000000 +0800 +++

After some parallel mount/umount test on ocfs2, we got this: slab error in kmem_cache_destroy(): cache `o2dlm_mle': Can't free all objects. Then we found a memleak situation in dlm_add_migration_mle(). When a mle found, it will be removed from dlm->hlist. If there is no pointer to it at that moment, the mle will become an ?orphan mle? that no process can find and release.