Niels de Carpentier wrote:
> I have a number of systems with an iscsi root filesystem. These systems
> connect to an redundant pair of iscsi servers, using tgtd. I use heartbeat
> to fail over the iscsi target. I'm using open-iscsi 869. It tried both the
> iscsi transport 869, and the default centos 724. The iscsid used was
> always 869.
> 
> I've set the replacement timeout high, so the iscsi root system should be
> able to recover from the short outage if the iscsi target fails over to
> another server:
> 
> node.session.timeo.replacement_timeout = 86400
> 
> Unfortunately, this doesn't always work. Sometimes the OS will report
> filesystem errors and mount the fs read-only. A short time later the iscsi
> targets will be reconnected, but the filesystem is already read-only by
> then.
> 
> The logs show (default iscsi transport 724 was used for this test):

Could you send the parts of the log before the fs errors? We want to see 
how many conn errors there were and when they occured and when the 
replacement/recovery timeout fired.

Could you also run
iscsiadm -m node -T target -p ip:port
for the root target and send the output?

And could you run

cat /sys/class/iscsi_session/session1/recovery_tmo

(session1 is the session for the root disk right. If not replace the one 
with whatever number it is).

and send that outpout?

> Apr 21 11:35:40 front003 iscsid: connection1:0 is operational after
> recovery (1 attempts)
> 

This is weird, because it only took one recovery attempt, so it looks 
like it was a really short outage and should have been back within the 
replacement you set.

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To post to this group, send email to open-iscsi@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/open-iscsi
-~----------~----~----~----~------~----~------~--~---

Reply via email to