These were accepted.
It was suggested, plausibly, that this might have been due to the recent change in the date format in the MAChistory to ISO 8601. Alastair will take a look. Action amended.
The batteries have been uplifted too now. ALL units should review any "to-be-kept" kit they have stored in there, with a view to getting rid of anything which can now be got rid of. Action tweaked.
Dave Berry had referred Craig to Miles MacCalman from IS, who had apparently been asking Schools (not including us) until he thought he had got enough information! Craig will pursue, and also make sure someone is at the technical group meeting on 19th August.
There was still no clear understanding as to what had caused this. Some people hadn't been affected at all; others had been with just about every machine they had upgraded. It was decided to defer the action to November. (RT#62244.)
Chris asked: "None of fs flush, fs flushmount or fs flushvolume makes the cache manager write data back to the server. Instead they clear the cache of things which are on the server. Would that still be useful?"
Chris will try asking on the OpenAFS users list to see whether it's really a problem, and if so what other folk do about it.
This part of the action is really complete. Fork a new action to cover emailing to ITPF.
Toby and Neil don't anticipate that non-ASCII names will cause problems but will try to fix if required.
Blog articles completed:
Report from Computing Executive Group
Some discussion followed Alastair's verbal report. He'll continue to try to get the details of this clarified.
A likely first step is for the search committee to try to put together a list of desirable attributes. It was thought best to have a group meeting with the CO/CSO community first, and then have individual meetings later if required.
A reminder to tell Alastair if you're interested in going.
Reports from units
Stephen reported that switching over to NFS/TCP had fixed the problem.
There was some discussion, but it was felt that the meeting wasn't quite the right place to thrash out the details. Ian will try it as an email discussion first, calling a separate meeting only if the email doesn't seem to be getting anywhere. Meantime, all were encouraged to read the previous email threads linked from the Unit report.
ALL to suggest websites (particularly non-inf.ed.ac.uk ones) to test.
Graham asked how the tests would be scheduled. Ian will pass on relevant details as he has them.
There was some discussion about recent installation problems, where updaterpms just appeared to lose contact part-way through. There wasn't a clear idea why, though the interaction of hardware bugs and the use of a mini-hub was suspected.
Topics for discussion
Alison asked ALL to re-read the wiki page, with a view to discussing at the next Operational meeting (14th August).
That's the plan!
The conclusion was these could mostly go now. Services-unit will take a look at what's there and whether they're still used, and get rid of any which aren't needed now.
Congratulations to Gillian and Richard!
Wednesday 14th August at 10:00 in IF-4.31
Please contact us with any comments or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh