[Linux-HA] advice on configuration

[George H]

Actually I had the same kind of problem as you. I never found any of
the official docs useful.
I managed to dig up 3 URLs that helpped me make my Gentoo HA nodes.

This is the gentoo tutorial I was talking about.
http://perlstalker.amigo.net/Gentoo/LVS.phtml

The rest are examples of better config files

http://www.howtoforge.com/high_availability_loadbalanced_apache_cluster
http://www.howtoforge.com/high_availability_nfs_drbd_heartbeat

To get my HA setup I had to read HA docs from the linux ha site going
all the way to Red-Hat and SuSE's docs on how to set things up. Not to
mention the MailTO script wouldn't work with me cos I was using
Postfix, had to write my own. lots of stuff was missing in the how-to
docs and I had to scavange for info in tons of mailing lists ..etc.
Then I had STONITH problems to which there was no way to see what
devices are supported except from looking at the actual stonith source
code.

Hope the links help, and google a lot and try to find any red-hat and
suse docs as well. They use pretty much the same software. Plus do a
search on this mailing list, I think someone mentioned something about
quorum.

On 1/12/07, Marshal Newrock <marshal at idealso.com> wrote:
> On Fri, 12 Jan 2007 07:30:29 +0200
> "George H" <george.dma at gmail.com> wrote:
>
> > This is the official Linux-HA Active-Active setup for heartbeat
> > http://linux-ha.org/GettingStartedV2/TwoApaches
>
> That does have a sample config, yes, but no information beyond the
> basic configuration.  For example, no mention of quorum or anything
> like that, or anything indicating that a simple two-node cluster will
> work as desired.
>
> > There is a better guide for setting up a multi-node active-active
> > heartbeat setup if you use gentoo, can be found on some gentoo-wiki
> > site.
>
> That only contains drbd setup right now.  I found drbd quite easy to
> configure.
>
> Given the complexity of heartbeat and my inability to even get it to
> start up, I've decided to instead use monit to monitor services (with
> the added benefit that it will also do "stand-alone" machines) and have
> it execute a script which causes fail-over when there's a service
> failure.
>
> --
> Marshal Newrock
> Ideal Solution, LLC - http://www.idealso.com
> _______________________________________________
> Linux-HA mailing list
> Linux-HA at lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>


-- 
"Nothing is impossibel for the person that doesn't have to do it"
"The probability of anything happening is in inverse ratio to its desirability"
"If I were a roman statue, I'd be made alabastard"
--
George H
george.dma at gmail.com