Difference between revisions of "Talk:Kit-retirement-autumn19"
m (→Old Webserver (Mantis))
(Add notice about uplift being arranged)
|Line 1:||Line 1:|
Revision as of 12:50, 22 October 2019
Created this page for people to discuss what the plan is with regards to retirement and potential disposal of some of the older gear in Tardis to make way for new things. As of 23/09/19 - the pile is looking as follows:
Servers in the pile to retire:
- BlueWhale (user:kline has confirmed they will pick this up when they are next in the country, although that may be some time)
- pangalacticgargleblaster (user:angusp)
- torchwood (user:angusp?) - probable raid card failure? (lots of IO timeouts)
- skaro (user:arkan)
- auton - raid array fucked
- gallifrey (user:arkan)
Very old things:
- Sunfire v210: ace
- nasty KVM
- gateway e-1400 pc
- dell optiplex pc of some p4-era vintage
- w2000 dell latitudes x2
- sun netra t1 (Mara)
- cisco PSU
- no-brand AMD semperon PC
- sun ultra-5 x2 (Leela and Gallifrey)
- 2x kontron telco servers (oracle and WOTAN)
- poweredge 1650 (piper)
- 3x poweredge 1850 (Ace, ace and Kal)
- hp dl380 gen4
- sunfire 280r (Mantis Broken (iirc old webserver? potentially with data that needs recovered?)
- poweredge 2950 ii
If anyone has any comments, lays claim to any of these machines or has any info that I'm not aware of regarding special significance and why they should be kept, these will likely be given to Informatics Student Support as they have said they will help us dispose of anything we want gone when they get CCL in to do the next recycling pickup. I'm going to set a provisional date of late October for this to happen.
I'm quite keen to keep a couple of Sun machines, just because they run SPARC and that's unusual, but also that's only if they are still functional. As for the PowerEdges, I can't see they have any particular significance?
Thought this would be easier to keep track of than an email chain, so please comment :D
RE the Netra
Also added some more links to the machines listed if they have wiki pages.
To clarify for future readers
It's being given to (Informatics) Computing Support, not Informatics Student Support (or ISS).
I have no particular attachment to any of the harderware in "Very old things". I know User:andrewferguson and User:camerongray wanted to hold onto something in that list, though. It's not clear as to whether you want to recycle bluewhale/torchwood/pangalagcdiddksdfkjsdf/gallifrey etc, but since the hardware is not ours, we probably should ask the owners first ! --Rphi (talk) 11:00, 30 September 2019 (BST) Qaisjp (talk) 16:31, 25 September 2019 (BST)
Yup, I'm keen to keep some of the interesting old hardware around. The netra is probably the easiest to work with, given that:
- It's supported by BSD
- It's rackmounted
- We (apparently) have the correct rails for it
- It's low-power at <100W
- It's 1 RU, and therefore doesn't take much space in the rack
As for the other stuff, well the Ultra 5's are neat. And one would probably go missing if it was destined for scrap. From recollection the remaining SunFire has been gutted for parts) [can someone confirm this?] and therefore isn't particularly useful.
Old Webserver (Mantis)
From discussions on IRC, this still has data on it and the "disks are borked". Things are complicated by the use of Fibre Channel disks, which I don't think are supported by any other machine we have. (can someone confirm this?). If that is the case, then it may be a good idea to keep Mantis until the data been recovered (or until it has been determined that a recovery will not be possible).
This is of course complicated by the issue that information about the issue with Mantis has been lost to time (or am I wrong? anyone have IRC logs?) and so it is not clear if the issue with the disks was caused by a hardware issue on Mantis (in which case keeping Mantis is probably not necessary).
Just to add, from some IRCing with Andrew, Mantis is the big Sun at the bottom of the pile, more details here: Archive:Mantis_Broken
If this is the same Mantis that I remember, there was a problem with the RAID array which either corrupted the disks or the array config. Not sure if I have IRC logs from the era but will take a dig. --Amazoph (talk) 23:33, 11 October 2019 (BST)