On 12/16/2011 07:57 PM, Emmanuel Dreyfus wrote:
Hi
I have two annoying bugs in 3.2.5. I wonder if there is a patch to
test, or a change to backport from master in order to address them
1) Adding a brick or rebalancing a volume changes inodes, and therefore
wreak processes that were working on f
On 12/17/2011 12:40 AM, Emmanuel Dreyfus wrote:
Pranith Kumar K wrote:
2) When using AFR, if a peer goes down, processes that have I/O pending
will se an error. Just retrying the same operation is fine, but that is
a bit furstrating.
Emmanuel,
Could you give the test case for the afr
On 03/13/2012 07:52 PM, Ian Latter wrote:
Hello,
Well we've been privy to our first true error in
Gluster now, and we're not sure of the cause.
The SaturnI machine with 1Gbyte of RAM was
exhausting its memory and crashing and we saw
core dumps on SaturnM and MMC. Replacing
the SaturnI h
On 03/13/2012 07:52 PM, Ian Latter wrote:
Hello,
Well we've been privy to our first true error in
Gluster now, and we're not sure of the cause.
The SaturnI machine with 1Gbyte of RAM was
exhausting its memory and crashing and we saw
core dumps on SaturnM and MMC. Replacing
the SaturnI h
number.
I didn't know this. Is there a queue length for what
is yet to be handled by the background self heal
count? If so, can it also be adjusted?
- Original Message -
From: "Pranith Kumar K"
To: "Ian Latter"
Subject: Re: [Gluster-devel] replicate background
On 01/09/2013 11:03 AM, Song wrote:
Hi,
We have a glusterfs clusters, version is 3.2.7. The volume info is as
below:
Volume Name: gfs1
Type: Distributed-Replicate
Status: Started
Number of Bricks: 94 x 3 = 282
Transport-type: tcp
We native mount the volume in all cluster servers. When w
On 01/07/2013 04:46 PM, Raghavendra Gowdappa wrote:
Pranith,
This comment is on the second patch. While the implementation looks fine, I've
some concerns related to the idea itself. Consider following situation with a
replicate volume of two subvolumes:
1. process 1 (p1) acquires a mandatory
hi,
Attaching the steps to re-create the issue.
As part of Entry transaction, before performing
create/mknod/mkdir/rmdir/unlink/link/symlink/rename fops, afr takes
appropriate entry locks and then performs pre-op. If the fop fails on
all nodes then the changelog leaves the directory in
On 01/16/2013 12:23 PM, huangql wrote:
Dear all,
I have encountered a strange problem that the ownership of
linkto file was changed to root:root after we reboot server. And I
reproduced the problem many times.
For example: ff_2
First, the new created file: ff_2
[root@test02 ~]# ll /test
On 01/23/2013 08:33 AM, Song wrote:
Hi,
When application access a directory of glusterfs volume, glusterfs
client spend about 1 hour, from 19:44:05 to 20:44:25.
The gluster volume is DHT + AFR, native mount to /xmail/gfs1
Access(/xmail/gfs1/xmail_dedup/gfs1_000/011/204/)
The following mess
On 01/09/2013 03:51 PM, Pranith Kumar K wrote:
On 01/07/2013 04:46 PM, Raghavendra Gowdappa wrote:
Pranith,
This comment is on the second patch. While the implementation looks
fine, I've some concerns related to the idea itself. Consider
following situation with a replicate volume o
hi Avati,
In entry self-heal of a directory, for every new file that is
created, data/metadata changelogs are marked on the source file to
indicate pending data/metadata self-heal from source file to the new
file that is created as part of this self-heal. But this data/metadata
changelog
On 02/01/2013 11:44 AM, Song wrote:
I have compared the source code of features/locks between 3.2.7 and
3.3.1. I find mutex lock is added in various functions, such as
get_domain function.
Are these changes updated to 3.2.7 branch?
*From:*Anand Avati [mailto:anand.av...@gmail.com]
*Sent:*
On 01/31/2013 10:14 AM, Pranith Kumar K wrote:
hi Avati,
In entry self-heal of a directory, for every new file that is
created, data/metadata changelogs are marked on the source file to
indicate pending data/metadata self-heal from source file to the new
file that is created as part of
hi,
Problem:
When there are multiple fds writing to same file with eager-lock enabled, the
fd which acquires the eager-lock waits for
post-op-delay secs before doing the unlock. Because of this all other fds
opened on the file face extra delay when
performing writes. Eager-locking, post-o
On 02/12/2013 08:43 AM, Anand Avati wrote:
On Mon, Feb 11, 2013 at 7:02 PM, Pranith Kumar K <mailto:pkara...@redhat.com>> wrote:
hi,
Problem:
When there are multiple fds writing to same file with eager-lock
enabled, the fd which acquires the eager-lock
hi,
Please note that this is a case in theory and I did not run into such
situation, but I feel it is important to address this.
Configuration with 'Eager-lock on" and "write-behind off" should not be
allowed as it leads to lock synchronization problems which lead to data
in-consistency among r
eager-lock code itself rather than depend on write-behind :|
Avati
On Mon, Feb 11, 2013 at 10:07 PM, Anand Avati <mailto:anand.av...@gmail.com>> wrote:
On Mon, Feb 11, 2013 at 9:32 PM, Pranith Kumar K
mailto:pkara...@redhat.com>> wrote:
hi,
Please note that
On 02/20/2013 07:03 AM, Anand Avati wrote:
On Tue, Feb 19, 2013 at 5:12 PM, Anand Avati <mailto:anand.av...@gmail.com>> wrote:
On Tue, Feb 19, 2013 at 3:59 AM, Pranith Kumar K
mailto:pkara...@redhat.com>> wrote:
On 02/19/2013 11:26 AM, Anand Avati wrote:
hi,
Could you let us know where is this feature used. I mean the use
cases it addresses.
Pranith.
___
Gluster-devel mailing list
Gluster-devel@nongnu.org
https://lists.nongnu.org/mailman/listinfo/gluster-devel
fs xlators. Is it ok if
we use xdata instead of flags to convey that write-behind took care of
overlaps?
Pranith
On Tue, Feb 19, 2013 at 7:20 PM, Anand Avati <mailto:anand.av...@gmail.com>> wrote:
On Tue, Feb 19, 2013 at 6:11 PM, Pranith Kumar K
mailto:pkara...@redhat.com>
On 02/23/2013 01:40 AM, Anand Avati wrote:
On Fri, Feb 22, 2013 at 5:39 AM, Pranith Kumar K <mailto:pkara...@redhat.com>> wrote:
hi,
Could you let us know where is this feature used. I mean the
use cases it addresses.
Can you be more specific about the question? N
hi,
From discussions with avati, we are thinking of addressing the
following issues in locks xlator.
At the moment to find which mount performed inode/entry lks,
the process needs to be attached to gdb and the connection address needs
to be de-referenced to figure out the actual
EXPECT "" get_dup_files $M0/dir6
EXPECT "" get_dup_files $M0/dir7
EXPECT "" get_dup_files $M0/dir8
EXPECT "" get_dup_files $M0/dir9
EXPECT "" get_dup_files $M0/dir10
Pranith.
>From 39b180fa54a2adbdf55d78454514b9a403a2
On 03/14/2013 10:57 AM, Emmanuel Dreyfus wrote:
With 3.4.0alpha2, ca processremains stuck on a FUSE operation on the
client. glusterfs client log says:
[2013-03-14 00:40:42.269307] W
[client-rpc-fops.c:1588:client3_3_finodelk_cbk]
0-gfs33-client-3: remote operation failed: Invalid argu
25 matches
Mail list logo