samuel
2013-Feb-12 18:19 UTC
[Gluster-users] server3_1-fops.c:1240 (Cannot allocate memory)
Hi folks,
We're using in an old environment gluster 3.2.2 and we've just started
seeing the following error in /usr/local/var/log/glusterfs/bricks/gfs-log:
[2013-02-12 19:15:05.214430] I [server3_1-fops.c:1240:server_writev_cbk]
0-cloud-server: 2400444243: WRITEV 7 (37067) ==> -1 (Cannot allocate memory)
[2013-02-12 19:15:19.463087] I [server3_1-fops.c:1240:server_writev_cbk]
0-cloud-server: 2249406582: WRITEV 15 (52493) ==> -1 (Cannot allocate
memory)
in a distributed replicated 2-node environment.
Both nodes have enough disk space (1.3T) but looks like used memory is
quite high:
node1:
total used free shared buffers cached
Mem: 4047680 4012700 34980 0 276 3728652
-/+ buffers/cache: 283772 3763908
Swap: 3906244 19916 3886328
node2:
total used free shared buffers cached
Mem: 4047680 4012488 35192 0 1088 3713512
-/+ buffers/cache: 297888 3749792
Swap: 3905532 25244 3880288
could it be a memory issue?
Best regards,
Samuel.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20130212/3e850c0e/attachment.html>
Brian Candler
2013-Feb-13 12:27 UTC
[Gluster-users] server3_1-fops.c:1240 (Cannot allocate memory)
On Tue, Feb 12, 2013 at 07:19:43PM +0100, samuel wrote:> Both nodes have enough disk space (1.3T) but looks like used memory is > quite high:Not true.> node1: > total used free shared buffers cached > Mem: 4047680 4012700 34980 0 276 3728652 > -/+ buffers/cache: 283772 3763908It shows: * total memory is 4GB * any spare memory is used to cache disk blocks. In this case it's 3.7GB. * the second line shows you the memory usage after disk buffers and cache are ignored. It shows you are only using 283M and have 3.7GB free. Ditto for the second box.
Joe Julian
2013-Feb-14 17:25 UTC
[Gluster-users] server3_1-fops.c:1240 (Cannot allocate memory)
Note that those are Info notices (" I "), not Errors (" E
").
3.2.2 has known critical bugs. You should upgrade (at least within the
3.2 branch) asap.
On 02/12/2013 10:19 AM, samuel wrote:> Hi folks,
>
> We're using in an old environment gluster 3.2.2 and we've just
started
> seeing the following error in /usr/local/var/log/glusterfs/bricks/gfs-log:
>
> [2013-02-12 19:15:05.214430] I
> [server3_1-fops.c:1240:server_writev_cbk] 0-cloud-server: 2400444243:
> WRITEV 7 (37067) ==> -1 (Cannot allocate memory)
> [2013-02-12 19:15:19.463087] I
> [server3_1-fops.c:1240:server_writev_cbk] 0-cloud-server: 2249406582:
> WRITEV 15 (52493) ==> -1 (Cannot allocate memory)
>
> in a distributed replicated 2-node environment.
>
> Both nodes have enough disk space (1.3T) but looks like used memory is
> quite high:
>
> node1:
> total used free shared buffers cached
> Mem: 4047680 4012700 34980 0 276 3728652
> -/+ buffers/cache: 283772 3763908
> Swap: 3906244 19916 3886328
>
> node2:
> total used free shared buffers cached
> Mem: 4047680 4012488 35192 0 1088 3713512
> -/+ buffers/cache: 297888 3749792
> Swap: 3905532 25244 3880288
>
> could it be a memory issue?
>
> Best regards,
> Samuel.
>
>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20130214/325a317c/attachment.html>