Quantum Forum for DXi V5000
DXi V1000 Troubleshooting
Hi All,
We had an unplanned power outage in our test lab, and our DXi V1000 has had problems since:
First boot -> home screen said "System State is Diagnostic". No healthcheck tab under utilities -> diagnostic. Left for a day or two, collected logs and rebooted.
Second boot -> home screen says "System State is Transient (Starting Services)". Again, No healthcheck tab under utilities -> diagnostic.
There is a service ticket for "DXi0: General DXi0 Software DataPath : Operation failure" with "ERROR: Service "DataPath" start failed with exit status 255!" - this is from prior to the second reboot.
My first question is; is there any way to log in to the DXi at the CLI? - it's CentOS under the hood and I'd love to be able to see what filesystems are mounted and what processes are running. The WebUI is sadly lacking in this regard.
Anyone have any ideas?
Many Thanks,
Barney.
Replies to This Discussion
-
Permalink Reply by Barney on
-
Update: After a couple of hours, "System State is Transient (Starting Services)" has become "System State is Diagnostic". A new service ticket for "DXi0: General DXi0 Software DataPath : Operation failure" with "ERROR: Service "DataPath" start failed with exit status 255!" has been generated.
It seems the only option in these circumstances is to log with Quantum support.
Would still like CLI access - anyone know if this is possible?
Cheers.
-
Permalink Reply by Jon McNerny on
-
Hi Barney,
There is some commands availible from the cliadmin. You can access this with cliadmin/cliadmin. Here attached is a guide on what you can accomplish with these credentals.
Although you cannot do alot of common commands from this shell, you can fine useful info from the collect log once gathered. This will show you /var/log/messages and other collected info (check out the collect.txt file). This might help troubleshooting.
Can you post the boot trace from the vm consule?
Jon
- Attachments:
-
-
2.x_CLIGuide_RevC_DXi-Series.pdf, 1.5 MB
-
-
Permalink Reply by Mark Thacker on
-
Hi Barney,
Yes, we do use CentOS with the DXi, though there is a lot of integration and value added software that makes up the whole DXi stack.
There is a fully documented and support CLI interface available through SSH. Login as "cliadmin" with password "cliadmin" and look at the "syscli --help" command. Additionally, the Command Line Interface Guide is available in the Resources Section on this Forum site. (see here : DXi V1000 Command Line User Guide)
It could be that the file systems themselves are simply not coming up. You should be able to determine a bit more through the CLI.
While we are continuing to investigate this on our end, it may be that more extensive support would require the use of Quantum Service. Such support is an option that customers may purchase for less than $100 annually per TB of vmPRO.
-
-
*Doh*
I meant that Silver or Gold Software support contracts are available for DXi V1000.
-
-
Hi Mark, and a Happy New Year,
Back after the festive break and had 5 mins to look at this - it looks like filesystems are not coming up - I have seen boot messages regarding vol0 being taken offline (I think).
As a lab environment, we're not really too fussed about rescuing data on this unit, but should we take this forward to production, we'd like to know:
1 - If this is typical behaviour? - if a DXi V1000 is uncleanly halted (e.g. if VMware HA pulled a DXi V1000 up on another host after host failure) does it generally come back OK - are there fsck's or similar that get performed? - have we been unlucky here?
2 - How this would be approached in a real-world scenario - is there a simple fsck / remount command we can run, or is this a Quantum Service case?
Finally, the linked doc. is not available with our serial number for some reason.
Many Thanks,
Barney.
-
-
Can you post the boot trace for us to review?
-
-
Hi Jon,
How can I best achieve this? - do you want syslog from the system diag file?
Many Thanks,
Barney.
-
-
Just copy and past what happens in the VM consule while booting please.
Jon
-
-
Hi Jon,
Didn't reboot, but hit the "Reset from Diagnostic State" - screen grab attached.
Many Thanks,
Barney.
-
-
Was this working and you had an unclean shutdown?
-
-
Yes - exactly. It was running fine, and then there was an unplanned power outage. It may well be that the underlying storage had some serious problems as a result - the battery cache on the RAID card was suspect. However, other VM's on the same host and datastore came back up OK, so we are curious to know what filesystem repair tools, if any, are available in such circumstances.
This is actually pretty critical, as it may inform infrastructure choices for production deployments.
Many Thanks.
-
-
-
I don't know why i can reply to your other Post Barney.
What does the GUI say your network setting are set to now?
Jon
© 2024 Created by Quantum Forum V.
Powered by
