Go to
previous page /
next page of Tier3 site log
29. 11. 2012 146GB disks to be absorbed by UIs
Unused Servers
server |
unused disks |
ex t3wn08 |
2 |
ex t3vmmaster01 |
4 |
t3dcachedb01 |
4 disks available but intouchables now, it hosts PNFS |
t3dcachedb03 |
4 if the VM t3dcachedb04 will proof to be ok as a Chimera node |
ex t3se01 |
2, but not now, we're still migrating |
RAID0 scenarios, OS and /scratch unsafe
Fabio's remarks
- to get 3 / 6 I can proceed next week.
- to get 4 / 6 I would expect Jan 2013.
- ok we can relocate disks but do we really want to stop to use those servers ?? new disks are cheaper than servers. Old servers can be turned in new UIs if we buy new disks.
Daniel's remarks
- /scratch space should only be used for temporary storage and not for valuable data, so I'm ok with the "unsafe" option but we should inform the users (I'm not sure all of them are aware of the current situation for t3fs02 and t3fs07)
- there is no rush to do this UI upgrade (other then the old lcg-utils but we can live with that for the moment), so why not wait until January if we can produce a more consistent setup then
- I will ask around for some use cases so have more input for the "safe /scratch" vs "large /scratch" decision
- even a large /scratch partition will be filled up within a few months; do we need an automated cleaning (also depends on the use cases of course)
Go to
previous page /
next page of Tier3 site log
This topic: CmsTier3
> WebHome >
CMSTier3Log > CMSTier3Log31
Topic revision: r2 - 2012-11-30 - DanielMeister