Thanks, it was that.
2015-03-16 17:47 GMT+01:00 Danny Auble :
> Yann, this is fixed in commit 2e2d924e3d0 and will be in 14.11.5 when
> released. The patch is quite small, all you will need to do it patch your
> code, recompile the mysql plugin and restart your slurmdbd.
>
> The problem comes f
08:22 AM, Yann Sagon wrote:
Re: [slurm-dev] Re: slurmdbd segmentation fault 14.11.4
#0 0x2b69ed3df395 in __strcasecmp_l_sse42 () from /lib64/libc.so.6
#1 0x2b69edc866de in as_mysql_node_down
(mysql_conn=0x2b6a04000960, node_ptr=0x2b6a000ffac0,
event_time=1426513631, reason=, reason_uid
#0 0x2b69ed3df395 in __strcasecmp_l_sse42 () from /lib64/libc.so.6
#1 0x2b69edc866de in as_mysql_node_down (mysql_conn=0x2b6a04000960,
node_ptr=0x2b6a000ffac0, event_time=1426513631, reason=, reason_uid=200) at as_mysql_cluster.c:1085
#2 0x00522c70 in clusteracct_storage_g_node_d
What is in the core dump back trace?
On March 16, 2015 7:30:07 AM PDT, Yann Sagon wrote:
>I have the same problem again after several days.
>
>slurmdbd -Dv
>
>[...]
>slurmdbd: debug2: Everything rolled up
>slurmdbd: debug4: got 0 commits
>slurmdbd: debug4: got 0 commits
>slurmdbd: debug2: Ope
I have the same problem again after several days.
slurmdbd -Dv
[...]
slurmdbd: debug2: Everything rolled up
slurmdbd: debug4: got 0 commits
slurmdbd: debug4: got 0 commits
slurmdbd: debug2: Opened connection 7 from 192.168.100.1
slurmdbd: debug: DBD_INIT: CLUSTER:baobab VERSION:7168 UID:200
I tried the hard way: revert back to previous slurmdbd, revert the mysql
backup, tried to remove some entries on the db. No luck.
Then I had a look at the dbd.messages in StateSaveLocation. I suppose it's
there where the messages to be flushed to the db are stored when the db is
not available.
I s
