Hi,
What can I do to maximize the performance of my openldap setup which running in master-slave mode using syncrepl? System has 8GM RAM and 4GB swap space Openldap-2.3.32 and DB.4.5 ---------------------------------- Total dn count is 2450000. LDIF size is 670MB using ldapsearch dbd size is : du -c -h *.bdb 69M authenticationId.bdb 227M cn.bdb 602M dn2id.bdb 27M entryCSN.bdb 24M entryUUID.bdb 2.5G id2entry.bdb 6.4M mail.bdb 4.6M objectClass.bdb 16M ou.bdb 263M sn.bdb 740M uid.bdb 4.4G total ----------------------- Portion of slapd.conf: cachesize 50000 checkpoint 4096 10 sizelimit 2500000 ------------------------- DB_CONFIG # Set the database in memory cache size. set_cachesize 2 0 1 # Set database flags. set_flags DB_TXN_NOSYNC set_lk_max_locks 2000 set_lk_max_objects 1000 set_lk_max_lockers 1000 set_txn_timeout 3000000 set_lock_timeout 3000000 set_lg_regionmax 1048576 set_lg_max 10485760 set_lg_bsize 2097152 set_lg_dir /usr/local/var/master/logs
db_stat -h /usr/local/master/db -m | head -n 2 2GB Total cache size 1 Number of caches
Thanks Pawan Kumar
--On Friday, November 30, 2012 1:09 PM +0530 Pawan Kamboj pawank.kamboj@gmail.com wrote:
Hi,
What can I do to maximize the performance of my openldap setup which running in master-slave mode using syncrepl? System has 8GM RAM and 4GB swap space Openldap-2.3.32 and DB.4.5
First off, upgrade to a supported release.
Total dn count is 2450000. LDIF size is 670MB using ldapsearch dbd size is : 4.4G total
So you should have a 2,450,000 cachesize setting. Your DB_CONFIG cachesize setting should be something like 4 0 1.
Or, migrate to using MDB in the latest openldap so you don't have to tune all BDB, and you get significantly improved perf.
--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
openldap-technical@openldap.org