A more detailed discussion about it: http://www.port389.org/docs/389ds/design/logging-performance-improvement.html

You could also disable logging and see whether the spikes disappear to be sure of their source: http://www.port389.org/docs/389ds/howto/howto-logsystemperf.html



Hi,

it could be flushing of logs (access.log) on disk which happens more often when server load is higher. you could use iostat or dstat to see what happens

Regards,
Andrey


De: "Ghiurea, Isabella" <Isabella.Ghiurea@nrc-cnrc.gc.ca>
À: 389-users@lists.fedoraproject.org
Envoyé: Mercredi 5 Septembre 2018 23:14:24
Objet: [389-users] ldap perfomance


Hello Gurus,

looking for an answer to  the following performance behavior

 my env: 389-ds-base-1.3.5.15-1.fc24.x86_64 in multimaster fractional replication

running rsearch for 5 min with 1 thread seeing spikes for a basic read  using index uid

And running with 10 threads  same search the  avg ms/ops  performance are much better with no major spike/burst

Any explanation much appreciate it


see bellow for 1 thread and the spike/burst

T 300 -t 1
rsearch: 1 threads launched.
T1 min=   0ms, max=   5ms, count = 54710
T1 min=   0ms, max=  42ms, count = 64930
T1 min=   0ms, max=   2ms, count = 65174
T1 min=   0ms, max=   2ms, count = 65110
T1 min=   0ms, max=  44ms, count = 64966
T1 min=   0ms, max=   1ms, count = 65101
T1 min=   0ms, max=  22ms, count = 65056
T1 min=   0ms, max=  32ms, count = 64981
T1 min=   0ms, max=   1ms, count = 65145
T1 min=   0ms, max=   1ms, count = 65223
T1 min=   0ms, max=  27ms, count = 65015
T1 min=   0ms, max=   1ms, count = 65182
T1 min=   0ms, max=   3ms, count = 65213
T1 min=   0ms, max=  23ms, count = 64760
T1 min=   0ms, max=   2ms, count = 64214
T1 min=   0ms, max=   3ms, count = 52279
T1 min=   0ms, max=  11ms, count = 64914
T1 min=   0ms, max=   1ms, count = 65118
T1 min=   0ms, max=   5ms, count = 64852
T1 min=   0ms, max=  91ms, count = 64180
T1 min=   0ms, max=   4ms, count = 64746
T1 min=   0ms, max=   1ms, count = 65080
T1 min=   0ms, max=  12ms, count = 65110
T1 min=   0ms, max= 702ms, count = 59243
T1 min=   0ms, max=   1ms, count = 65082
T1 min=   0ms, max=  89ms, count = 64331
T1 min=   0ms, max=  23ms, count = 64647
T1 min=   0ms, max=   5ms, count = 64818
T1 min=   0ms, max=  55ms, count = 64374
T1 min=   0ms, max=   8ms, count = 64713

T1 min=   0ms, max=   8ms, count = 64713
300 sec >= 300
Final Average rate: 6394.22/sec = 0.1564msec/op, total: 64713


 And final avg rate for 10 threads, no significant spike/burst  for this num of threads


20180905 14:07:23 - Rate: 16962.10/thr (16962.10/sec = 0.0590ms/op), total:169621 (10 thr)
300 sec >= 300
Final Average rate: 17420.40/sec = 0.0574msec/op, total:169621




_______________________________________________
389-users mailing list -- 389-users@lists.fedoraproject.org
To unsubscribe send an email to 389-users-leave@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/389-users@lists.fedoraproject.org

_______________________________________________
389-users mailing list -- 389-users@lists.fedoraproject.org
To unsubscribe send an email to 389-users-leave@lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/389-users@lists.fedoraproject.org