Strange problems with heartbeat....
alanr at unix.sh
Sat Aug 25 07:06:01 MDT 2001
The best way to ask these kinds of questions is to send them to the linux-ha
mailing list. I've CCed the mailing list in my response.
Pawel Kisiel wrote:
> I have the following problem with heartbeat:
> I have 2 firewall boxes configured with heartbeat...and almost
> works great...except one little thing....
> When I turn off the primary HA machine the secondary takes all
> resources of the primary....and when the first comes up it should give up
> all of them...but it doesn't do that :(
> The funny thing is that the communication on serial heartbeat is
> OK because in logs of primary machine I can see that the secondary box is
> active and so on....
> I had no problems when I was testing this configuration with only
> one IP address to manage....but I'm using the Linuxvirtualserver and I
> need several IP address on world interface...
People do this quite commonly.
> I don't understand why if this situation comes up, I can log on
> first box using static IP addres given on world interface to maintain
> administrative jobs...but when the resources on second box are not
> released(even if heartbeat is stopped on the second one) I don;t have even
> ping communication to this machine??
> What I figured out, using first box is that I can see MAC address
> of the second one using arp -a ip_of_the_second_box so the machine and
> its world interface is still alive, but I cannot log on to this machine
> or even to ping it.....
> In this situation, when I reboot the second machine everything is
> OK and I can do everything with it...
> In logs of the secondary firewall I can see that heartbeat can see
> that primary box is ok and heartbeat is releasing IP resources but it is
> only in logs...ifconfig shows me that IP are still not released.....
> Can U help me with that problem?
OK. It is possible that there is a bug in ifconfig on your machine - or
that something else is wrong. Could you please reply to this note and
attach both logs from both the primary and secondary machines?
> And one question in my present situation.. I can't get in to the
> second box but first is alive and I know that static ip for ssh connection
> is UP( ping ip_of_second_box and arp -a ip_of_second_box shows me its MAC)
> and I know that Heartbeat package is stopped on it so the only one IP
> is up righ now....
> What can I do ? I tried to do arp -d on the ip of second box...
> I don't have the direct access to the second one to reboot it
> right now, so do I have to wait till Monday when I'll be at work?
I don't know. Does anyone else on the list have a suggestion?
> BTW:Thank You for help with compiling the heartbeat....
Glad to have been of help. Hope we can help you this time too ;-)
-- Alan Robertson
alanr at unix.sh
More information about the Linux-HA