Datacentre tasks
From Wikitech
(Redirected from Datacenter tasks)
DO NOT ADD THINGS TO THIS PAGE. PUT THINGS IN RT.WIKIMEDIA.ORG. DO NOT DELETE ANYTHING ON HERE YET, ROBH WILL TAKE CARE OF IT.
Anyone adding to this page:
Please add things to the first section "Unsorted New Tasks." This way Rob will make sure he is aware of the task before moving it to a more specific heading. This will ensure nothing gets dismissed by having a high priority task mixed in with a lot of low priority items. As always, if there is any question, do not hesitate to contact Rob directly.
Unsorted New Tasks
- Hosts which are down since the power outage: db19, storage2, storage3, isidore. -- Tim 23:07, 5 July 2010 (UTC)
Outstanding Sun Issues
- db28 has more dimm errors --RobH 18:48, 4 March 2010 (UTC)
- Was in node 1 dimm 6 & 7, moved them to node 0 1 & 2 and dmesg sill shows a ton of memory errors for those dimms.
- sun case 72708580
- swapped the bad memory out, still bad. Bad memory was in 0 1:2 and is now sitting in DC, but still has errors on box:
[ 40.475014] K8 ECC error.
[ 40.506322] EDAC amd64 MC1: CE ERROR_ADDRESS= 0x813dcb630
[ 41.580861] Northbridge Error, node 1
- Errors occur both before and after memory swap.
- db34 LOM not working
- in rt.wikimedia.org
- db16 in rt.wikimedia.org
(the below need checking and possible removal) --RobH 13:52, 24 May 2010 (UTC)
- db13 has failed RAID controller battery: http://p.defau.lt/?XsaDOS3KdZHeCO9VzypUnw - needs replacement --Domas Mituzas 12:46, 21 May 2009 (UTC)
- db14 RAID battery is dead
- Sun Serial # 0820QAS047
- db18 is throwing raid controller errors.
- Sun Case # 71443454
- db26 is having the same controller issues as db18.
- ms3 no LEDs lit, but is online
- Sun SN: 0850AMR088
- Resetting the system via LOM doesn't clear this issue. Firmware upgrade may, but will require downtime. --RobH 16:10, 23 February 2010 (UTC)
