|Computing Meeting February 20 2008 [message #5930]
||Wed, 20 February 2008 11:37
Registered: June 2004
Location: GSI, Darmstadt
a) setup of test cluster for global redirector with Fabrizio|
- status: up and working, but:
- currently only sequential file copy is possible
- several bugs were detected and submitted
- procedure: bugs were submitted to Andy. As soon as they are fixed Fabrizio provides new tarball which will be installed at GSI.
already now the cluster can be used to copy files from Castor to GSI:
procedure (not official way, though).
</sim/2007/LHC07c/pp_minbias/6000/006/ >whereis root_archive.zip
Feb 20 11:30:29 info The file LHC07c/pp_minbias/6000/006/root_archive.zip is in
SE => ALICE::CERN::CASTOR2 pfn => root://lxgate38.cern.ch:1094//castor/cern.ch/alice/2005_castor2/07/54219 /dd33f5fc-2d28-11dc-afa9-000e0c3e6d91.1183881444
xrdcp -f root://lxts07.gsi.de:1094//alien/05/27473/e7e519e4-2c93-11dc-878d-kschwa rz@lxg0503:/tmp$ xrdcp root://lxts07.gsi.de:1094//alien/07/54219/dd33f5fc-2d28-11dc-afa9-000e0c 3e6d91.1183881444 /tmp/root_archive.zip
XrdClientConn::GoToAnotherServer: log id: 0, streamid: 1
Disabling apmon monitoring since env variable APMON_CONFIG was not found
[xrootd] Total 61.16 MB |====================| 100.00 % [9.4 Mb/s]
now the file can be accessed on local disk or at lxts07 cluster as if they were local at GSI.
kschwarz@lxg0503:/tmp$ unzip root_archive.zip
b) File transfer via AliEn
Currently transfer efficiency: 39%. 92 successful transfers today. A further problem occured. The queue was stuck when too many files were requested which were not staged yet. Fixed yesterday night by Andreas Joachim Peters. So far no transfer completed from our list of transfers.
c) lxfs47 back
the machine which was burned has been put back to our xrd cluster
d) xrd works now with gStore, seemingly reliable. The gstore backend is being tested on the lxfs50 cluster
e) Software: all ROOT, AliRoot and AliEn versions + software dependencies will be supported on Debian Sarge (32) and Debian Etch (32 and 64). Currently several updates are in progress and the login scripts are reorganised. Result will be published on Wiki.
f) currently is being investigated how many machines we can safely operate in RZ2. The issues concerning elecricity and cooling are being optimised under the given circumstances.
[Updated on: Wed, 20 February 2008 12:10]
Report message to a moderator