[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
Re: Failover Master setup
- To: Digant C Kasundra <digant@uta.edu>
- Subject: Re: Failover Master setup
- From: matthew sporleder <msporleder@gmail.com>
- Date: Fri, 24 Jun 2005 11:23:25 -0400
- Cc: Openldap-Software <openldap-software@OpenLDAP.org>
- Content-disposition: inline
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=sRTXLMp5chBihLgQUDYqzJNhFiM7rj/5pOKEGHdfm+i6zfdCUfzrkT6A95EPI8xTS88WJZhKqWcrFFc3T03iTRl2SL7QUbAwjxzB3I+2CODeU9wxpombbBL5kwcNGVfEvTkNF/efdUQufyTDrieT00pSyDpFDYz7gabg1OxZVBo=
- In-reply-to: <1119545536.18838.18.camel@localizer.uta.edu>
- References: <1119545536.18838.18.camel@localizer.uta.edu>
Use Veritas to float an IP and mount.
It will just move the whole app to a different server on failure.
If you want protection against data-inconsistancy/corruption on an
active-active multi-master setup, then I would be interested to see
how other LDAP's implement ACID-ity. :)
On 6/23/05, Digant C Kasundra <digant@uta.edu> wrote:
> Hello everyone,
>
> A month or two down the line, we are hoping to have our PeopleSoft
> Student Information System write directly to our LDAP directory. As
> such, the question of "why are we using OpenLDAP if it can't do
> multi-master" is cropping up from my superiors. As OpenLDAP does not
> support multi-master (for good reason, I think), I was wanting to try to
> figure out a way to setup a failover master and was wondering if anyone
> has done work in this area?
>
> As it would work in my head, M1 would replicate to M2 and all the
> slaves. A heartbeat on M2 would monitory to make sure M1 was active.
> If M1 goes down, M2 would rewrite its config to replicate to M1 and the
> slaves and would take over as the master server.
>
> Anyone given thought to this sort of setup? I've never had my master
> server fail but I need to appease the naysayers. :)
>
> -- DK