Blog articles suggested:
Blog articles completed:
Report from Computing Executive Group
Reports from units.
The action and deferred action relating to the AT server room were taken with this item.
A number of the machines on the lists which were circulated to the COs list are scheduled to be retired, so it appears as though we should not be quite as tight for space as we had thought. However all should keep up the pressure to have unneeded machines removed.
Ian was able to remind us that the deferred action referred to the idea of putting cold spares on Dexion shelving in AT, in case there were problems in the Forum which might require their use. The general issue of hot and cold spares in AT will be deferred until the end of next month, by which time Alastair should have a better idea of how the hand-down process is working.
Now that we have enough power provision, the next problem is likely to be a shortage of rack space. Dave H has confirmed that there is space for at least one more rack, and possibly two (though the second less conveniently positioned). Research staff have been polled for their plans for the room, though with little response as yet.
If any machines in the room are actually managed by us, it would be useful to have them moved through to the main server room to free space. All to consider.
As an aside, the switches in the Forum are now four years old, and it's time we considered replacing them. George and Alastair to discuss.
Chris noted that DICE_STICK_WITH_SL62 had been applied only to <stable> machines, so if anyone wants their <develop> machines held back they would need to add the definition to their profiles themselves.
Followup from Chris in the chat room: "My 6.3 machine is on 2.6.32-279.5.1.el6 so we have the fix for 834764. We don't yet have the fix for 841983 so it's probably best, for now, to keep any machines using vlans on a bonded interface at SL6.2 (since 841983 doesn't mention bnx2)."
Alastair noted that the IBM array's firmware is desparately in need of an upgrade. Once that is done, metropolitan can be redeployed as a kvm host.
Toby remarked on the difficulty of getting information out of IS, citing the recent issue of the non-creation of some new-student mail accounts as an example. He'll provide details to Alastair off-line.
There was some discussion of the issues with printing slightly non-standard things to the new MFDs, and concern was expressed that these had not been picked up during the evaluation phase. Schools (certainly Informatics) had not had much time to exercise them thoroughly, and so had had to rely on the central evaluation. Craig will raise the issues with the upstream printing/MFD team, with a view to getting a better test suite in place for the next time.
Topics for discussion
Only inf-unit's ARP protection article, as noted in their report
The next meeting will be on Wednesday 26th September, 10am in 4.31. The convener will be George.
Please contact us with any comments or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh