SR failed to complete operation-> LUN not found

Printable View

11-21-2011, 01:25 PM
fritecsas

SR failed to complete operation-> LUN not found

Helo
I have set up a dss6lite system and created the Volumes etc. I also created the icsi targets, logical volumes etc, one time with Block I/O one time using File I/O. I always used the given namings just to do a basic setup.
However, no chance to get the SR connected to XenServer 5.6SP2. The target IQN Scan was ok, the Target LUN Scan results in the error message, that SR failed to complete the operation.
I am somehow confused.

Can anybody get some light into this matter?
thanks a lot.
11-21-2011, 02:54 PM
Al-S

How much the size of your LUN ? is it more than 2TB ?

Also did you tried to quit your ISCSI session and reconnect?
11-21-2011, 03:43 PM
fritecsas

Helo
the size is 1800 gb plus the overhead for replication etc.

I did reconnect iscsi sesions, I also did restart both servers (dss server and Xenserver), no changes.

iscsi_discovery (iscsi-ip) did work ok on the xen cli.

the discovery of the LUN results in the mentioned error. The logfile from the xen /var/log/auditlog does give some wird information

Nov 21 16:37:36 PCCO-XEN01 xapi: [20111121T15:37:36.503Z| info|PCCO-XEN01|3922|Async.SR.create R:742206fc7782|audit] ('trackid=717c3b128093ff2e954947c0d9bd241a' 'LOCAL_SUPERUSER' 'root' 'ALLOWED' 'ERROR:SR_BACKEND_FAILURE_96: [ ; The request is missing or has an incorrect target IQN parameter; <?xml version=\"1.0\" ?> <iscsi-target-iqns> <TGT> <Index> 0 </Index> <IPAddress> 10.33.2.12 </IPAddress> <TargetIQN> iqn.2011-11:dss.target0 </TargetIQN> </TGT> <TGT> <Index> 1 </Index> <IPAddress> 10.33.2.12 </IPAddress> <TargetIQN> * </TargetIQN> </TGT> </iscsi-target-iqns> ]' 'API' 'SR.create' (('host' 'PCCO-XEN01' 'fd04bab9-d233-40b4-8d43-e3baa25be4f5' 'OpaqueRef:06b501b6-8830-5dac-4a69-89b22b96ba44') ('name_label' '__gui__' '' '') ('name_description' 'SHOULD NEVER BE CREATED' '' '')))

If this would be the first open-e iscsi, I would probably skip this experience an go back to local disks :rolleyes:

thanks
11-21-2011, 04:11 PM
fritecsas

additionals:

I have probbed the SR using CLI commands. here are the results: I am not able to receive the SCSIid back from the Target.

[root@PCCO-XEN01 ~]#
[root@PCCO-XEN01 ~]# xe sr-probe type=lvmoiscsi device-config:target=10.33.2.12 device-config:targetIQN=iqn.2011-11:dss.target0
There was an SR backend failure.
status: non-zero exit
stdout:
stderr: Traceback (most recent call last):
File "/opt/xensource/sm/LVMoISCSISR", line 378, in ?
SRCommand.run(LVHDoISCSISR, DRIVER_INFO)
File "/opt/xensource/sm/SRCommand.py", line 244, in run
sr = driver(cmd, cmd.sr_uuid)
File "/opt/xensource/sm/SR.py", line 128, in __init__
self.load(sr_uuid)
File "/opt/xensource/sm/LVMoISCSISR", line 75, in load
iscsi = driver(self.original_srcmd, sr_uuid)
File "/opt/xensource/sm/SR.py", line 128, in __init__
self.load(sr_uuid)
File "/opt/xensource/sm/util.py", line 1135, in transform
return func(inst, *args, **kwargs)
File "/opt/xensource/sm/ISCSISR.py", line 167, in load
self._initPaths()
File "/opt/xensource/sm/ISCSISR.py", line 170, in _initPaths
self._init_adapters()
File "/opt/xensource/sm/ISCSISR.py", line 236, in _init_adapters
self.devs = scsiutil.cacheSCSIidentifiers()
File "/opt/xensource/sm/scsiutil.py", line 129, in cacheSCSIidentifiers
line = "NONE %s %s %s %s 0 %s" % \
IndexError: list index out of range
11-21-2011, 04:29 PM
Pi-L

You have to forgive me as I am not a XEN expert here, but here is some information about this error you can find on the net:

http://old-list-archives.xen.org/arc.../msg00158.html
11-21-2011, 06:06 PM
fritecsas

Thanks for the link. I tried this patch, by then it did not even do the target scan.
I re installed the dss6-lite plus created all volumes new.
no chance.

Looks like, I have a problem using dss6-lite with Citrix Xen 5.6SP2
11-21-2011, 06:11 PM
Pi-L

Have you tried other iSCSI Targets? I am just wondering how can we help you if this looks like Xen issue.
11-21-2011, 06:11 PM
Gr-R

can you show us your default initiator settings from Xen ??
11-22-2011, 07:32 AM
fritecsas

Hi Pi-L
this is the problem, from this location, I have no other iSCSI Target available, the productiv iSCSI Targets are all in other locations separated in VLANS.
I am trying bto figure out something, to get another Target available. Thought about the same thing yesterday. unfortunately, the Buffalo I am having here for NAS does not do iSCSI, only NFS.
I also thought of this might be an XEN issue, however, have some trouble in fihuring out the real sourc of problems. Both systems are new and have never prouven to be ok.
The XEN works with the NFS Share and the local store, The iSCSI open-e is ready prepared with the 1,8 TB Volume.
11-22-2011, 08:25 AM
red

try mabye something like:
iqn.2011-11.dss.target0 or iqn.2011-11.dss:target0 as a name
11-22-2011, 08:48 AM
fritecsas

now you got me: where to check the default initiator setting? I tried to find the info in connection to the iscsi and see the connection status at the open-e status info page:

I did a manuel discover and connect
[root@PCCO-XEN01 ~]# iscsiadm -m discovery -t st -p 10.33.2.12
10.33.2.12:3260,1 iqn.2011-11:dss-lite.target0

then logged myself in
[root@PCCO-XEN01 ~]# iscsiadm -m node -L all
Logging in to [iface: default, target: iqn.2011-11:dss-lite.target0, portal: 10.33.2.12,3260]
Login to [iface: default, target: iqn.2011-11:dss-lite.target0, portal: 10.33.2.12,3260]: successfu

The open-e showed the following connection:
Initiator name: iqn.2011-11.com.example:7beceb9f

After logging off, I tried to connect to the target
[root@PCCO-XEN01 ~]# iscsiadm -m node -T iqn.2011-11:dss-lite.target0 -p 10.33.2.12 -l
Logging in to [iface: default, target: iqn.2011-11:dss-lite.target0, portal: 10.33.2.12,3260]
Login to [iface: default, target: iqn.2011-11:dss-lite.target0, portal: 10.33.2.12,3260]: successfull

another probe at the XEN results in the same error:

[root@PCCO-XEN01 ~]# xe sr-probe type=lvmoiscsi device-config:target=10.33.2.12 device-config:targetIQN=iqn.2011-11:dss-lite.target0
There was an SR backend failure.
status: non-zero exit
stdout:
stderr: Traceback (most recent call last):
File "/opt/xensource/sm/LVMoISCSISR", line 378, in ?
SRCommand.run(LVHDoISCSISR, DRIVER_INFO)
File "/opt/xensource/sm/SRCommand.py", line 244, in run
sr = driver(cmd, cmd.sr_uuid)
File "/opt/xensource/sm/SR.py", line 128, in __init__
self.load(sr_uuid)
File "/opt/xensource/sm/LVMoISCSISR", line 75, in load
iscsi = driver(self.original_srcmd, sr_uuid)
File "/opt/xensource/sm/SR.py", line 128, in __init__
self.load(sr_uuid)
File "/opt/xensource/sm/util.py", line 1135, in transform
return func(inst, *args, **kwargs)
File "/opt/xensource/sm/ISCSISR.py", line 167, in load
self._initPaths()
File "/opt/xensource/sm/ISCSISR.py", line 170, in _initPaths
self._init_adapters()
File "/opt/xensource/sm/ISCSISR.py", line 236, in _init_adapters
self.devs = scsiutil.cacheSCSIidentifiers()
File "/opt/xensource/sm/scsiutil.py", line 129, in cacheSCSIidentifiers
line = "NONE %s %s %s %s 0 %s" % \
IndexError: list index out of range

I will see if I can get me some infos from the XEN Forums.
Thanks for your help.
11-22-2011, 11:13 AM
red

i meant the target name, not initiatior
11-22-2011, 11:21 AM
Pi-L

If you still cant tell what is malfunctioning the ultimate test would be connect to iSCSI target for example from Windows initiator, if it works that would mean the rest is up to XEN configuration. Have you tried that?
11-22-2011, 11:31 AM
fritecsas

Hi Pi-L
I thought of that. I will get me a W2k8R2 server this afternoon to check the connection to the open-e.
I was curious to do so in the first time, since the w2k8 would be a vm on that XEN SRV. I am getting me a W2k8 labserver put on another xen, which has good iscsi plugs to a standard open-e V6.

Keep you postetd.
11-22-2011, 01:01 PM
fritecsas

Hi Pi-L
looks like it is a Problem of the installed Xen.
I used another XENSERVER, which is just for test and lab pourpose, but basicaly not in the same IP segmnet. Putting it into an IP range which could be reached via WAN, Icould plug the open-e dss6-lite.

Now I need to find out whats wrong with this XEN Server, since this is the same installation build as the one which could do the plugging
Thanks for your help.
11-22-2011, 01:04 PM
Pi-L

If it is not in production yet always best way to start is to clean everything and install again.. :)
11-22-2011, 03:13 PM
fritecsas

Hi
found the solution:
I had an USB Device plugged on the Xenserver to import the vm files. Even though, it was no longer mounted, this IOMEGA HD blocked the cach of the scsiid which caused the plug from the iscsi volume to fail.

http://forums.citrix.com/thread.jspa...rt=15&tstart=0

Shoud throw it into the garbage.
01-17-2012, 06:24 AM
Jemshope

I also thought of this might be an XEN issue, however, have some trouble in fihuring out the real sourc of problems. Both systems are new and have never prouven to be ok.