| Computing Meeting, Wednesday, January 23, 2008 [message #5683] |
Tue, 22 January 2008 12:02  |
Silvia Masciocchi
Messages: 162
Registered: May 2006
|
first-grade participant |
From: *gsi.de
|
|
Meeting at 12:30 in room 4.161
Minutes by Victor Penso
Preliminary Agenda:
-------------------
1. Status of transfer of December data
2. Report on xrootd developments with Fabrizio
3. Data storage and management at GSI
? priviledges for individual users implemented?
? news about the fuse interface?
? when usable space in lustre?
4. GRID analysis jobs (Ana, Kilian)
5. monitoring of the batch machines and report of failures
[Updated on: Wed, 23 January 2008 11:09] Report message to a moderator |
|
|
|
| Re: Computing Meeting, Wednesday, January 23, 2008 [message #5743 is a reply to message #5683] |
Wed, 30 January 2008 10:03  |
Kilian Schwarz
Messages: 91
Registered: June 2004
Location: GSI, Darmstadt
|
continuous participant |
From: *gsi.de
|
|
Minutes:
1. Status of transfer of December data
so far 0% to GSI.
Current status:
- Files that already made it to another T1 (FZK in most cases) are advertised with the FTS protocol and because GSI doesn't have FTS those transfers fail
- There was and still seems to be something wrong with the Transfer Optimizer, there are still messages like:
Jan 18 10:10:04 info The transfer failed due to: contacting the Broker/Transfer
- Pablo is looking into these.
2. xrootd developments with Fabrizio
The newest xrootd software packaged by Fabrizio has been installed on lxfs60, lxfs61 (data servers). Redirector is planned on lxts07 (not accessible currently).
Design: no MSS backend for test cluster planned since data to be copied are anyway a backup from somewhere else.
3. data storage and management at GSI
- priviledges for individual users: implemented via Anna by using the xrd shell and ROOT. Open for testing.
- news about fuse: parts of fuse still do not install. Asked Andreas Petzold and Fabrizio for help.
- when usable space for lustre: HA solution for central DB server and DB backup not tested so far. If Lustre is used by ALICE then people should be aware of the risks. PBM will ask Walter for time scale.
4. monitoring of batch machines and failures:
Friday Super Micro and CPI will visit GSI. From the last 25 boxes we bought for D-Grid and ALICE roughly 1/3 stopped working within the first 6 months. This will be discussed with the Super Micro representatives.
New GSIAF monitoring link:
http://lxgrid3.gsi.de:8080/stats?page=CLUSTER/table
5. GRID analysis jobs:
they work now. 32bit and 64bit and various compiler versions have been decoupled. But to be able to read the old data for analysis they have to be moved from grid1 and grid2 in the AliEn FC.
|
|
|
|
GSI Forum
Search
Help
Members
Login
Home
