Hello,

I'm using DRBD with heartbeat (R1 config) and ipfail.

Here are my versions
CentOS 5.3
heartbeat-2.1.3-3.el5.centos
drbd-km-2.6.18_128.1.14.el5-8.3.1-3
drbd-8.3.1-3 (compiled from source and installed with rpm)

Each node (mail1 and mail2 ) has two interfaces.

eth0 -> heartbeat link, vip and uplink
eth1 -> hearbeat link and DRBD replication link

To ensure a failover if one nic fails, i set ipfail to ping the gateway.

First i testet my configuration with a firewall-rule to trigger ipfail.

# iptables -A OUTPUT -p icmp --icmp-type 8 -j DROP

That worked perfectly well.
You can see it in the attached files messages.mail1.iptables and
messages.mail2.iptables.
I takes over the vip and drbddisk starts without an error.

But if i try to pull the cable (or ip link set eth0 down) on mail1
eth0, the drbddisk resource doesn't get stopped on mail1
and so it fails to start on mail2.
You can see that in messages.mail1.linkdown and messages.mail2.linkdown.

Jul 23 15:51:12 mail2 ResourceManager[11462]: [11696]: ERROR: Return
code 1 from /etc/ha.d/resource.d/drbddisk
Jul 23 15:51:12 mail2 ResourceManager[11462]: [11697]: CRIT: Giving up
resources due to failure of drbddisk::home

You can also see these files attached:
ha.cf
haresources
drbd.conf

I don't rellay know if it's a problem with heartbeat or drbd, but i
hope you can give me a hint.

Thank you for your help!

Attachment: drbd.conf
Description: Binary data

Attachment: ha.cf
Description: Binary data

Attachment: haresources
Description: Binary data

Attachment: messages.mail1.iptables
Description: Binary data

Attachment: messages.mail1.linkdown
Description: Binary data

Attachment: messages.mail2.iptables
Description: Binary data

Attachment: messages.mail2.linkdown
Description: Binary data

_______________________________________________
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