[Linux-HA] R1 to R2 testing: cib.xml & ldirectord questions for 2 node cluster

Peter Farrell peter.d.farrell at gmail.com
Thu Sep 13 06:32:54 MDT 2007


Hi Andreas -

Everything seems fine. I responded to Andrew in relation to what I'd
asked Dejan.
(Regarding failover w/ no network connectivity)

We use the R1 version of heartbeat / ldirectord as follows:

On our DMZ segment:
2 nodes (running hearbeat w/ ldirectord) connected via serial
2 webservers running apache w/ content on NFS

One of the 2 nodes says 'I have addresses 192.168.1.1-10'
(which map to either of the 2 webservers via ldirectord)

Each of the 2 webservers would have 10 loopbacks (lo:0 - lo:9)
configured for that address range.

Each unique request that arrives should be sent to each webserver in
turn via round robin using ldirectord.

The idea was that I'd get redundancy in the cluster w/ heartbeat and
load balancing in the web servers via ldirectord. In the real world it
works most of the time - sometimes the round-robin doesn't work at all
or one of the nodes will go right up it's backside - but generally
it's pretty effective.

-Peter

On 13/09/2007, matilda matilda <matilda at grandel.de> wrote:
> >>> "Peter Farrell" <peter.d.farrell at gmail.com> 12.09.2007 17:24 >>>
> > Andreas -
> >
> > A follow up if you will...
>
> Hi Peter,
>
> do you need further assistence. I saw that Dejan helped you.
> Is everything fine and running?
>
> In which mode does you use ldirectord? (NAT?)
>
> Best regards
> Andreas Mock
>
>
> _______________________________________________
> 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
>



More information about the Linux-HA mailing list