On Sep 27, 2012, at 4:42 PM, iscsi developer man iscsidevel...@gmail.com
wrote:
2.6.32-37-generic
if scsi commands are timing out, we would expect to see TASK ABORTS, correct?
Yes. Are you taking a tcpdump trace at this time and seeing task aborts?
With the 2.6.32 based kernel setup
we are using ubuntu 10.04 12,04 and centos 5.3
Mike, in some cases we will see the conn errors even without any nop timeouts.
From what I've seen, the initiator stops sending NOPS when the
connection is doing I/O. (I will confirm in our logs today)
I saw a patch elsewhere on this mailing list
On 09/26/2012 11:55 AM, iscsi developer man wrote:
we are using ubuntu 10.04 12,04 and centos 5.3
Mike, in some cases we will see the conn errors even without any nop timeouts.
What kernel version is ubunutu using?
It could be scsi commands are timting out? Do you get 1011 errors or
1021
we are building an iscsi-target in our lab here, that consist of many
nodes across multiple racks..
we are running multiple clients to this cluster, and although we are
not having issues with our I/O, open-iscsi is repeatedly outputting to
/var/log/messages the following two logs.
19:12:04.703
On 09/25/2012 03:45 PM, iscsi developer man wrote:
we are building an iscsi-target in our lab here, that consist of many
nodes across multiple racks..
we are running multiple clients to this cluster, and although we are
not having issues with our I/O, open-iscsi is repeatedly outputting to