These were accepted as a reasonable account of the last meeting.
An initial page is here, but still needs some work done on it.
There had been some confusion as to what this referred to, but it was clarified that item 3 of the inf-unit report for 28th September was what was meant. It was suggested that dropping it into RT (RT#79713) would ensure it wouldn't get lost.
It works for Toby, but there's a command-line step involved. Ditto for MIT kerberos. These are straightforward, but it still might be worth wrapping them in some kind of clickable thing. Action passed to Services-unit.
cosign will be done as part of its SL7 upgrade.
Defer to January
Craig has had some helpful discussions with users. He'll be writing up his conclusions in due course.
Alastair has been fleshing out the points made in a way more suitable for general consumption, but reckoned there was no reason he couldn't just circulate the prioritised list now. Since done.
Actions added or revived:
Neil had remembered what the issues were, and had emailed to the cos list about them. If there's to be more than one virtual spindle then discs which are not already part of a RAID set should be marked as RAID-ready.
Blog articles suggested:
Report from Computing Executive Group
It had been IS's suggestion to wire PPLS back to level 2. We reckon that this'll make future use of space less flexible. For discussion elsewhere.
Reports from Units
It was agreed that for anything (not just switch firmware upgrades) which might have a global impact, even though it wasn't expect to be visible at all, the procedure should be to email to the cos list a few minutes in advance and then not proceed until that email had been seen looped back to the sender. That way we could be reasonably sure that the warning would have made it out to anyone who might need to know.
The Forum core diagram could do with its labels being updated for the 10GbaseT change. (Done. Note that this diagram doesn't show the srNN copper links, only the fibre links to the closet switches.)
ALL to ensure that these are done, including any followup reboots for kernel upgrades.
Tim and Alison will arrange for the support team to be shown how the new facility works. We do expect support to be done mostly by others.
There had been multiple causes for user account issues recently, which has meant that each has had to be investigated separately. For the PGT->PGR rollover, upstream changes in the procedure from ?Registry had meant that some student settings were being retained inappropriately. There had also been (different) problems with new staff accounts.
Tim and User Support will be meeting to review procedures.
Tim will write a blog article and probably also send out a general email.
There is a meeting scheduled (Thursday 27th) to discuss the theon/prometheus interface.
Topics for discussion
Alastair explained the proposals for Appleton Tower level 7. There followed an extensive discussion of the merits and demerits of moving all of the Computing team to ATL7. All to let Alastair know ASAP about any additional points, as the decision has to be made very soon.
Do remember to check the menus and fill out the doodle poll. Details are here.
The next meeting will be on Wednesday 26th October 2016 at 10:00 in IF-4.31, chaired by George.
Please contact us with any comments or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh