[Date Prev][Date Next]
RE: Master replication server reliability
- To: "Buchan Milne" <firstname.lastname@example.org>
- Subject: RE: Master replication server reliability
- From: "James Courtney" <Jcourtney@inphonic.com>
- Date: Thu, 12 Aug 2004 12:52:12 -0400
- Cc: "Open LDAP Software List \(E-mail\)" <openldap-software@OpenLDAP.org>
- Content-class: urn:content-classes:message
- Thread-index: AcSAjAAAfU7xvCvdTSuBplnPjFHN4AAAG0NA
- Thread-topic: Master replication server reliability
I actually do have checkpoints set at 32k/60 mins.
As mentioned, we do really VERY few updates (perhaps one every few minutes) as the only reasons for updates are the following.
1) new user
2) change of password
3) change of first/last name
Do my checkpoint values seem okay?
I assume I should use the same checkpoint values on the slave servers too right?
From: Buchan Milne [mailto:email@example.com]
Sent: Thursday, August 12, 2004 9:49 AM
To: James Courtney
Cc: Open LDAP Software List (E-mail)
Subject: Re: Master replication server reliability
-----BEGIN PGP SIGNED MESSAGE-----
James Courtney wrote:
| I will try to build/install 2.2 in our dev environment today. I believe
| I've seen some emails on this list WRT fast loading using slapadd if
| certain transaction parameters are set/disabled somehow in Berkeley DB.
| Is this a practical/safe thing to do for data loading to a new server if
| one re-enables things correctly when done? Is this a standard practice
| documented somewhere that I could educate myself on further?
Please try setting the "checkpoint" attribute for your bdb database
first ... the symptoms you describe and your log extract seem to
indicate you have no checkpoint setting ...
Buchan Milne Senior Support Technician
Obsidian Systems http://www.obsidian.co.za
B.Eng RHCE (803004789010797)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
-----END PGP SIGNATURE-----