Help needed

[Alex Kramarov]

I had such a situation once (and i think that this is not the case here, but
anyway) :

machine a : 200.0.0.1/25
machine b : 200.0.0.2/25
shared ip  : 200.0.0.3/25

heartbeat on machine a saw heartbeat on b, but b didn't see the heartbeat on
a and was always becoming primary. the machines were pinging each other and
the config files were perfect. the problem was that machine b was configured
with netmask 255.255.255.0 and not 255.255.255.128

(one more time i saw such an event was when i had iptables blocking inoming
udp on one of the machines).

maybe someone can write a "heartbeat ping" utility (like ping, or tnsping in
oracle - just an application level ping), just to check if the machines talk
to each other on the network level, in the same way heartbeat does.

----- Original Message -----
From: "Juri Haberland" <haberland@altus.de>
To: "Nicolas Bouthors" <nicolas@bouthors.org>
Cc: <linux-ha@muc.de>
Sent: Monday, January 14, 2002 3:43 PM
Subject: Re: Help needed


> Nicolas Bouthors wrote:
>
> > Hi,
> >
> > I'm quite new to heartbeat, and I can't manage to make it work correctly
> > for my setup :
> >
> > I'm trying to build a two-node cluster with high availability (drbd +
> > heartbeat). My problems are :
> >
> > When one of the servers boots, it systematicly takes the resouces (ip +
> > services) from the other node (that was running primary). How can I tell
> > heartbeat to start in secondary mode except if it's "alone" ?
>
>
> Put 'nice_failback on' into your ha.cf file.
>
> > In my setup, when I shutdown -h the primary active machine, heartbeat on
> > the the other one says 'other-node is dead' but doesn't take up
> > resouces.
> >
> > Those two problems prevent me from doying anything usefull whith
> > heartbeat. Any idea where I did something wrong ?
>
>
> Do you have the node names in ha.cf (remember: it must match `uname -n`):
> node host1
> node host2
>
> Do you have one (or more) lines in the haresources file like this:
> host1 aaa.bbb.ccc.ddd service1 service2
>
> where aa.bbb.ccc.ddd is the _shared_ IP address of your cluster?
>
> Are you sure that you didn't configure the shared IP address on any host?
>
> Juri
>
> --
>    If each of us have one object, and we exchange them,
>       then each of us still has one object.
>    If each of us have one idea,   and we exchange them,
>       then each of us now has two ideas.
>
>
> --------------------------------------------------------------------------
----
> Linux HA Web Site:
>   http://linux-ha.org/
> Linux HA HOWTO:
>
http://metalab.unc.edu/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html
> --------------------------------------------------------------------------
----
>
>
> .