Hi David,

Thank you for the explanation you gave about anycast.
Heartbeat doesn't seem to implement this feature.
Do i have to conclude that it is simply impossible to build a cluster on two
different subnets with Heartbeat?

Regards,

Smain


Selon David Lee <[EMAIL PROTECTED]>:

> On Wed, 5 Nov 2008, Matthew Soffen wrote:
>
> > On Wed, 2008-11-05 at 07:54 +0100, Smain KAHLOUCH wrote:
> >
> >> Hi,
> >>
> >> Do somebody know how to make heartbeat work when two nodes are on two
> >> different subnets ?
> >> Do you need more information ?
> >>
> >> Thanks,
> >>
> >> Smain
> >>> Hi all !
> >>>
> >>> I was wondering if there is a simple way to configure heartbeat in order
> to work
> >>> across the network.
> >>>
> >>> I mean, the two nodes are on two different subnets. (OpenLDAP Cluster).
> >>>
> >>> I looked for information about that. I just found the following topic but
> the
> >>> method seems to be complicated.
> >>>
> >>> http://www.gossamer-threads.com/lists/linuxha/users/31977#31977
> >>>
> >>> Thank you for your help :)
> >>>
> >>> Smain
> > [...]
> > Hi Smain,
> >
> > Are you really sure that is what you want to do ?
> >
> > How are you going to have them share a single IP address (the cluster
> > address) if it isn't portable across the 2 subnets ?
>
> The single IP address might be capable of being portable, by using
> "anycast".
>
> Disclaimer: I have no experience whatsoever of "anycast".  But it does
> seem to be there, and in real use.
>
> Although I've been using IP networking for years, that term "anycast" was
> new to me just last year.  But apparently it, too, has been around for
> years (but presumably very quiet).  And it is even, apparently, used for
> some of the DNS root servers, giving a single IP address multiple-instance
> presence on different continents.  Or something like that.
>
> But "anycast" is not so much heartbeat-like failover; rather multiple
> running instances, with clients accessing their nearest instance, as
> determined by routing deep in the network.  (Your client accesses one
> instance; mine accesses a different one; yet both (physical) instances
> offer the same (virtual) IP address because the routing supports it.)
> And if a particular instance failed, the routing would simply find another
> instance of that same IP address.  (Or, again, something like that.)
>
>
> --
>
> :  David Lee                                I.T. Service          :
> :  Senior Systems Programmer                Computer Centre       :
> :  UNIX Team Leader                         Durham University     :
> :                                           South Road            :
> :  http://www.dur.ac.uk/t.d.lee/            Durham DH1 3LE        :
> :  Phone: +44 191 334 2752                  U.K.                  :
> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>


_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to