Currently the gpfs xrootd space is handled by a single xrootd server that connects directly to the redirectors. The list below shows how to move the server from one gpfs machine to another one without causing clients to fail (there will be a short stalling of the clients).
Before Moving
- remove restart from taylor on <old> (wasn't even enabled)
- Update crontab on <new>
- add <new> to xrootd.cf to distribute proper xrootd file
- backup xrootd.cf on <old> backup/xrootd.cf
- edit (in place) xrootd.cf on <old> add:
export /fgstdummy/ if cmsd
Restart
- activate prod -> v4.1.1/ on <new>
- Stop cmsd on <old>
RDR will suspend writes and reads for files not found. Xrootd will continue to run and clients can read/write.
- [ ] Start xrootd/cmsd on gpfs05
files on gpfs will be accessible
- [ ] Start cmsd on gpfs02
- RDR will reconfigure gpfs02 and remove all old entries
- [ ] If no more clients connected to gpfs02 stop xrootd/cmsd
- [ ] rm prod link fermi-gpfs02
prod -> v4.1.1/
- [ ] let Charlotte know that Nagios needs to be changed
- [ ] change glast,glastall host lists
*** After [0/1]
- [ ] enable restart in taylor on gpfs05
Overview
Content Tools