Hello Mike,

I've responded inline.


On Fri, Apr 11, 2014 at 12:50 PM, Mike Christie <[email protected]>wrote:

> On 04/03/2014 09:32 AM, Rushikesh Jadhav wrote:
> > Hello Forum,
> >
> > We have a IBM V3700 and 4 Xenserver6.2 hosts. We are facing trouble in
> > adding SR from it and below are the logs generated in kern.log
>
> What is "SR"?
>
>
SR stands for storage repository. It simply means logging on to the target
LUN from host and creating a LVM layer on it.

>
> > n2 kernel: [429671.643262]  connection19:0: pdu (op 0x2c itt 0x1)
> > rejected. Reason code 0x7
> > n2 kernel: [429671.643302]  connection19:0: pdu (op 0x2a itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643354]  connection19:0: pdu (op 0x2b itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643524]  connection19:0: pdu (op 0x2f itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643554]  connection19:0: pdu (op 0x2e itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643703]  connection19:0: pdu (op 0x2d itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643750]  connection19:0: pdu (op 0x33 itt 0x1)
> > rejected. Reason code 0x4
> > n2 kernel: [429671.643846]  connection19:0: pdu (op 0x35 itt 0x1)
> > rejected. Reason code 0x7
>
> The target does not like something that the initiator side is doing.
> Probably, something in xen is sending a command that the target does not
> support. Can you log into the target and access some of its logs? There
> should be something about a protocol error.
>
> Could you take a wireshark/tcpdump trace?
>

The target is IBM-V3700 storage. We do not have access to its controller
logs. After contacting IBM & giving them the controller dump, they said
"everything is fine on controller and there are no errors".

I can take a tcpdump from host, could you please let me know the exact
options ?
Usually we do it with # tcpdump -v -i eth5


> >
> > Unfortunately VMs running from existing iSCSI luns face IO timeout as
> > the LUNs on host hang. This hang occurs as soon as the new LUN is
> > getting added.
>
> What do you mean by added? Do you mean you export them on the target
> side then discovery them on the initiator side? Or on the initiator at
> the iscsi/scsi level do you see them and then are you adding them to
> some xen management type of application?
>

I mean, the LUN is exported from target and then discovered on host. While
its getting mounted, suddenly the iSCSI driver faces a hang with kernel
messages showed above. This leads lockup on all mounted LUNs for 5mins till
the iscsi initiator recovers from the error automatically.


>
> >
> > Existing luns can be recovered only if we do restart iscsid and get no
> > success in adding additional luns.
> >
> > The iscsid is running with default parameters.
> >
> > The iscsi module losses all LUN access even for non-IBM as soon as it
>
> What other targets?
>

Other targets such as software iscsi or dell equallogic or EMC.


>
> > tries to mount LUN from IBM-V3700. The iscsi rpm
> > is open-iscsi-2.0.871-0.20.3.xs1120
>
> That is no help. I have no idea what xen puts in their iscsi code on top
> of upstream.
>
> Ok. I can provide src.rpm as well.

>
> Could you send the logs? Does xen store the kernel logs in
> /var/log/messages? Could you send that for when this happens?
>

The logs are stored in /var/log/messages and are same as mentioned earlier.
n2 kernel: [429671.643262]  connection19:0: pdu (op 0x2c itt 0x1) rejected.
Reason code 0x7
n2 kernel: [429671.643302]  connection19:0: pdu (op 0x2a itt 0x1) rejected.
Reason code 0x4


Let me know if you need any other logs.

-- 
You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/open-iscsi.
For more options, visit https://groups.google.com/d/optout.

Reply via email to