On Sat, Jul 30, 2011 at 6:31 PM, Quanah Gibson-Mount quanah@zimbra.com wrote:

--On Saturday, July 30, 2011 6:23 PM -0300 Friedrich Locke friedrich.locke@gmail.com wrote:

Please don't top post.

Sorry.

...

It is not that's using 63MB. It is that is uses 63MB after performing 4 lookup, i.e., its memory usage grows from 15MB to 63MB.

...

cachesize       4096

How many entries do you have in your database?

No more than 50 entries at most.

...

dbnosync dirtyread

These are *not* good values to set.  You should remove them.

I will remove then.

You have failed to provide some key pieces of information.  Please provide your DB_CONFIG file.  Please provide the total size of your BDB database (du -c -h *.bdb).  Please provide the number of real cores your system has available.  ITS6660 *only* affects systems with 4+ CPUs.  All I've seen you say so far is that as slapd is used, it grows in size.  That's typical of slapd loading entries from off of disk into memory.  Unless we know how large your database itself *is*, there is no telling if what it is doing is wrong or not.

My DB_CONFIG is:

set_cachesize 0 16777216 1 set_lg_regionmax 262144 set_lg_bsize 2097152

My database size is:

gustav# du -c -h *.bdb 32.0K cn.bdb 32.0K dn2id.bdb 64.0K id2entry.bdb 32.0K mail.bdb 32.0K mailAlternateAddress.bdb 32.0K mailHost.bdb 32.0K objectClass.bdb 32.0K uid.bdb 288K total gustav#

My system has 2 cores!

--Quanah

--

Quanah Gibson-Mount Sr. Member of Technical Staff Zimbra, Inc A Division of VMware, Inc.

---

Zimbra ::  the leader in open source messaging and collaboration