Thanks, a lot everyone!
Performance looks good now. No deadlocks and nothing being purged from the cache. I have started pointing more sssd clients at the cluster.
Thanks,
Douglas Duckworth, MSc, LFCS HPC System Administrator Scientific Computing Unit Physiology and Biophysics Weill Cornell Medicine E: doug@med.cornell.edu O: 212-746-6305 F: 212-746-8690
On Thu, Aug 31, 2017 at 2:51 AM, Ulrich Windl < Ulrich.Windl@rz.uni-regensburg.de> wrote:
Douglas Duckworth dod2014@med.cornell.edu schrieb am 30.08.2017 um
16:06 in Nachricht CAAKHBKm0FKBR4BLNpy6jrszFp35Eid_J2F19MKmfS+G47EKCJA@mail.gmail.com: [...]
Only four clients are currently using this cluster so perhaps I should actually use DB_CONFIG before putting it into production.
Maybe you could start considering this from a moderately loaded and sized server: set_cachesize 0 15000000 1 set_lg_regionmax 262144 set_lg_bsize 2097152 set_flags DB_LOG_AUTOREMOVE set_lk_max_locks 30000 set_lk_max_objects 30000
Also note "olcDbConfig: <DB_CONFIG setting>" is probably the modern way to do it.
Regards, Ulrich