These were accepted as a reasonable account of the last meeting.
Folk were reminded that the review of pages with a review date prior to mid-2014 should be completed by the beginning of April. (The first Ops meeting in April is Wed 13th.)
CEG-owned pages should have a corresponding action there.
The January (sic) release is due, so Chris will look to producing a set of diffs when that one is made.
Actions added or revived:
Report from Computing Executive Group
Suggested topics for the meeting:
Please send any further suggestions to Alastair ASAP.
Alastair gave a verbal report on progress, and the possible options for and consequent on level 6.
Reports from Units
George will recreate the list and circulate.
It was clarified that this referred to both external non-University users as well as those from other Schools; and that our users on non-Informatics subnets (such as at South College Street) were also included. We want to get this in as part of the ongoing EdLAN review, so comments and suggestions to George ASAP please.
There was a power outage for (some of) Wilkie Building this morning. E&B report that a "transformer tripped" and that there were "widespread outages in medics area". There'll be further work carried out in the area tomorrow (Thursday 10th) evening.
It wasn't clear how this would affect the entropy pool.
Neil wondered whether there might be at least one missing "dcs-style" X-related package.
[Post meeting: the packages involved are dice-fvwm-startup, dice-xdm and dice-dcs-Xstartup. For SL6 they're all in core/packages/dice/dice_sl6_unsupported.rpms, while for SL7 the first and third are in core/packages/dice/dice_sl7_env_graphical.rpms while dice-xdm doesn't appear to be in anything at all. dice-xdm does have some scripts which are useful for hooking into session-managers other than xdm, in addition to the component itself.]
The KVM server need some firmware upgrades. This will be scheduled over the next few weeks, and will entail some downtime and/or temporary migrations.
There should really be better monitoring of the spool space.
There was some discussion as to the best way to expand this. While there were several schemes proposed, it seemed as though reinstalling might be the easiest.
Could we impose size limits? "It depends." There are ways in which a small submission may be expanded by cups as a temporary file along the way. And what happens when lots of students all try to print simultaneously in one of the labs?
It seems really that having a decent amount of spool space, which is monitored, is the way to go.
Topics for discussion
This was held over.
The next meeting will be on 9th March 2016 at 10:00 in IF-4.31, chaired by Craig.
Please contact us with any comments or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh