A bit of background on this problem. We've got two sparc Solaris 9 Sunfire V480Rs in a Veritas cluster with the cluster managed by VCS. The cluster offers nfs services to a large number of workstations. When i benchmarked the i/o speeds to the cluster, i found that the read bandwidth was high, while the write bandwidth was atrocious. Turns out the problem was a veritas limitation caused by the number of nfsd processes that VCS launches. VCS starts only 16 nfsd processes, irrelevant of the value of NFSD_SERVERS in /etc/default/nfs. Took us awhile to find that one. _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagersReceived on Thu Feb 12 08:59:10 2004
This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:43:29 EST