[Linux-HA] Heartbeat on Fedora Core 3

Jason Whiteaker jwhiteak at verizon.net
Thu Mar 17 17:10:13 MST 2005


I have built Heartbeat on FC3 (v 1.99.3), and I'm having some success. 
However, there's an issue I can't figure out. I can fail over a test 
apache service several times, and things seem to work OK. Log entries 
look good. However, after about 7 minutes, I receive the following 
message in the ha-log:

heartbeat[5101]: 2005/03/17_18:37:03 info: all clients are now paused

Also, when I try to issue a "/etc/init.d/heartbeat stop" on the backup 
node then main node, the shutdown process will often get hosed. I 
sometimes receive this error:

heartbeat[2153]: 2005/03/17_18:53:11 WARN: glib: TTY write timeout on 
[/dev/ttyS0] (no connection or bad cable? [see documentation]).

It would seem that the ttyport is having some issue, but this is a known 
good, commercial null-modem cable. I'm running all ha.cf options at 
their default (e.g. baud=19200, etc.).

When I used 1.99.2, the above problem happened all of the time. Now, 
with 1.99.3, it seems to be a little more stable, however, we're still 
seeing this flaky "all clients are now paused" thing on both nodes.

Any input is welcomed! I'm not a programmer, so please bear with me! :-)


More information about the Linux-HA mailing list