RE: slapd memory and performance problems

[Quanah Gibson-Mount <quanah@stanford.edu>]

--On Wednesday, April 02, 2003 9:13 AM -0800 Mike Denka <mdenk@whidbey.net> 
wrote:

> I'm thinking now after reading yours and Matthew Backes' explanation of
> memory utilization, and also since doubling the server RAM this morning
> to 1GB and discovering that the performance problem still exists, that
> this is not a memory problem but some other issue for which I yet have
> no explanation.  CPU utilization is very low.  Disk writes are numerous
> to the var partition where the ldap log file and the openldap database
> are, but reads to that partition are very low. So am I correct in
> assuming that much data is cached (thus the high memory utilization by
> the threads)?  Then might the problem lie in the network connections?  I
> am currently seeing right around 100 connections, most in TIME_WAIT.  Is
> this something that could be corrected, at least in part, by unix
> sockets (ldapi:// url)?  If so, can you point me to docs describing its
> use?
>
> Quanah, in another response to this thread suggested the possibility of
> RedHat threading issues.  Have you or anyone else run into any problems
> with RedHat's threading implementation?

I did with RedHat 7.2, and also with Solaris 9.  Which why we deployed on 
Solaris 8.  I have not tried RedHat since that time.  I'm waiting for our 
software group here to get me compiled versions of Heimdal for Linux before 
I move to testing it again.  In a conversation I had with Howard, he did 
say there can be issues using glibc & gcc less than 3.x.  So if you are 
using an older gcc to compile your binaries, that could be a problem.

We ourselves are not using ldapi for our sendmail connections, so far as I 
know, and when I do my load tests, I'm simply using ldapsearch against the 
host from 18+ remote hosts (I've gone as high as 70), with consistent 
positive performance.

One question is, where do you store your log files?  We had an enormous 
boost in write efficiency, when per a suggestion by Howard, we moved our 
bdb log files to a different partition than the db directory.  It cut the 
load time for our LDIF files from 3 hours to 1 hour.  If you are seeing a 
lot of writes to your system, that could impact performance (although I 
doubt it would be this much).

--Quanah

--
Quanah Gibson-Mount
Senior Systems Administrator
ITSS/TSS/Computing Systems
Stanford University
GnuPG Public Key: http://www.stanford.edu/~quanah/pgp.html