Dear,
After some rigorous checks & debug found the local FS was getting exhausted due to huge amount of logs written by slapd daemon.
Unfortunately, this FS is system mounted locally (sda2) & hence couldn't be expanded further. As a workaround, has restricted log quantity to 6000/10 seconds via rsyslog rate-limiting feature.
Ulrich, Will check onto system for any slow local disk problem persists.. Many thanks for your kind & valuable suggestion. ----
*Thanks & Kind Regards,* Saurabh LAHOTI.
On Thu, 28 Jun 2018 at 09:17, Ulrich Windl < Ulrich.Windl@rz.uni-regensburg.de> wrote:
Saurabh Lahoti saurabh.astronomy@gmail.com schrieb am 27.06.2018 um
23:44 in Nachricht CAB5-MqpjqiX-PLYmV64fj5o-EYA4dtC4+w0U8d=7NL+ndeByxg@mail.gmail.com:
Dear,
We've started observing strange & peculiar performance behavior for slapd instances. While ldap bind response is more than 1.5 second on one server then on another it's just 0.006 seconds.
Both the servers have similar OS, system resources & ldap version.
OS: RHEL 6.9 LDAP: 2.4.44 Backend: bdb
Could you please suggest some probable causes..?
Could it be a slow local disk?
Thanks & Kind Regards, Saurabh Lahoti.