Dataflow machine down

[Anusha Ranganathan]

Hello Sander,
The dataflow website has gone down again and from the error message
looks like it needs yet another file system check.
Did you get an account and are you able to access the VSphere management
client. I can access the VSphere client at bourgeois.oerc, but I am
still unable to access the console and so can't initiate the fsck. I can
restart the VM, but from past experience, this doesn't help.

If you do not have access to the client, no worries. I have opened a
ticket with OERC. May be this time I'll get lucky and someone will act
on the ticket :)

Regards,
Anusha

> -----Original Message-----
> From: Sander van der Waal [mailto:sander.v...@oucs.ox.ac.uk]
> Sent: 22 June 2012 12:47
> To: sup...@oerc.ox.ac.uk
> Cc: Anusha Ranganathan; Katherine Fletcher
> Subject: RE: [rt.oerc.ox.ac.uk #16629] Dataflow machine down
>
> Thanks Luke! I'm sitting at a talk so can only quickly check
> the end result and the website has come up again so it looks
> like it's fixed. Many thanks for your help.
> If you can create an account for me that'd be great.
> My Oxford ID is oucs0096 should you need that.
>
> Kind Regards
>
> Sander
>
> OSS Watch - supporting open source in education and research
> http://www.oss-watch.ac.uk
>
>
> > -----Original Message-----
> > From: Luke Raimbach via RT [mailto:sup...@oerc.ox.ac.uk]
> > Sent: 22 June 2012 12:45
> > To: Sander van der Waal
> > Subject: RE: [rt.oerc.ox.ac.uk #16629] Dataflow machine down
> >
> > Dear Sander,
> >
> > I connected to the console of dataflow.oerc and saw it was waiting
at a fsck
> > prompt for someone to press 'F'.
> >
> > I pressed 'F' and fsck continued and freed about 4 orphaned inodes.
The
> > machine then rebooted and looks to have come up normally. Are you
able to
> > verify this?
> >
> > Also, you don't have an account on our Active Directory, so wouldn't
be able
> > to connect to the vSphere management machine at
haute-bourgeois.oerc. Would
> > you like me to create you an account?
> >
> > Thanks,
> > Luke.
> >
> > > -----Original Message-----
> > > From: Sander van der Waal via RT [mailto:sup...@oerc.ox.ac.uk]
> > > Sent: 22 June 2012 11:51
> > > Subject: [rt.oerc.ox.ac.uk #16629] Dataflow machine down
> > >
> > >
> > > Fri Jun 22 11:51:21 2012: Request 16629 was acted upon.
> > > Transaction: Ticket created by sander.v...@oucs.ox.ac.uk
> > > Queue: OeRC-Support
> > > Subject: Dataflow machine down
> > > Owner: Nobody
> > > Requestors: sander.v...@oucs.ox.ac.uk
> > > Status: new
> > > Ticket <URL:
https://rt.oerc.ox.ac.uk/Ticket/Display.html?id=16629 >
> > >
> > >
> > > Dear Luke,
> > >
> > > Our dataflow.ox.ac.uk website has been down for a while.
> > > I believe it's hosted on the dataflow.oerc.ox.ac.uk machine,
managed by
> > > OeRC. I've just been brought in to take a look at this but I can't
connect
> > to
> > > that machine from the University network via SSH, is that a policy
issue?
> > > There's a virtual desktop account for admin on the machine via
haute-
> > > bourgeois.oerc.ox.ac.uk but my credentials for user volt0032 are
not
> > > working either.
> > >
> > > This has been looked at before and the website has been down for a
> > > number of days now, so can this please be looked at as a matter of
priority?
> > >
> > > Many thanks,
> > >
> > > Sander
> > >
> > > OSS Watch - supporting open source in education and research
> > > http://www.oss-watch.ac.uk
> > >
> > >
> >

[Alexander Dutton]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Anusha and Sander,

The people at the OeRC sent this earlier:

> Below is a list of VMs which were running on one of our Data
> Stores which experienced a very long I/O lag during a failover
> event yesterday afternoon.
>
> The Data Store did not timeout so no problems were immediately
> apparent. However, it was later discovered that some of our VMs
> that run on this Data Store did not like the long wait for I/O.
>
> A small handful of our own VMs exhibited the following behaviour:
> - BSD systems panicked and rebooted after SCSI timeouts (some did
> not recover gracefully) - Linux systems re-mounted their root file
> systems read-only (rebooting these systems restored them to normal
> service) - Windows systems just waited and recovered, no reboot
> (no intervention was necessary)
>
> I'm sorry to have to bring you this bad news. If your VM is in the
> list below, please log in via the vShpere Client on
> haute-bourgeois.oerc.ox.ac.uk and check on the status of your
> machine(s):
>
> dataflow.oerc.ox.ac.uk,Powered On,Normal
> dataflow-test.oerc.ox.ac.uk,Powered On,Normal
> dataflow-vm1.oerc.ox.ac.uk,Powered On,Normal
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJP8swJAAoJEPotabD1ANF7i6YH/RPxmPnmV7g0VygHUyb289i3
+cOW8UbnnVRm6FDTcO7aJdMTuEnQMAU6RVxXMGPJ1yb3xz/xlkHXDMNBgQxJBKR1
4wXz87sE14ljRuoyCD1nDOoVnmqF+/62rtsyk+/bRu7NvbtLLzZqJZeoroR68aRH
IC38tVze6yQyYZkESbqpU/KI52VmHrCGwmPfFpqhO5Lxsla915qX5LAuMTNli1dX
pkJ/pn0XJlFrXazvZ2treRcwhV/UfKBy1A8eh6mo0LKfThGSY6/BF/3Y2bIAt00p
5ipRoMimKjztDqIt8gPVurvT+YEXnlGHPcHm9lT8Rj+JW7uUmX5s4VnOSBGZeaA=
=1t+4
-----END PGP SIGNATURE-----

[Ross Gardler]

I believe that one of the points I made about using a foundation for
hosting software and related services is that we no longer have to
worry about maintaining our own services in this way.

just saying ;-)

Ross

On 3 July 2012 11:26, Anusha Ranganathan

> --
> Post to: dataflo...@googlegroups.com
> Unsubscribe: dataflow-deve...@googlegroups.com
> Web: http://www.dataflow.ox.ac.uk/
> http://groups.google.com/group/dataflow-devel?hl=en
> https://github.com/dataflow



--
Ross Gardler (@rgardler)
Programme Leader (Open Development)
OpenDirective http://opendirective.com