[Linux-HA] base64_to_binary: input length invalid

[Luettgen, John (N-ENSCO)]

Please help to understand ... I have an active/passive node configuration, using sles 9, kernel 2.6, heartbeat 2.0.3.  The node became unresponsive and was forced to poweroff each, the only clue is the subject message that appears in the log nearly 200 times.  The ha.cf is also included below.

Ideas anyone?

10:52:22 uspstms1 logd: [9264]: info: logd started with default configuration.
10:52:22 uspstms1 logd: [9264]: WARN: Core dumps could be lost if multiple dumps occur
10:52:22 uspstms1 logd: [9264]: WARN: Consider setting /proc/sys/kernel/core_uses_pid (or equivalent) to 1 for maximum supportability
10:52:22 uspstms1 logd: [9274]: info: G_main_add_SignalHandler: Added signal handler for signal 15
10:52:22 uspstms1 logd: [9264]: info: G_main_add_SignalHandler: Added signal handler for signal 15
10:53:23 uspstms1 ipfail: [9494]: debug: PID=9494
10:53:23 uspstms1 ipfail: [9494]: debug: Signing in with heartbeat
10:53:23 uspstms1 ipfail: [9494]: debug: [We are uspstms1]
10:53:23 uspstms1 ipfail: [9494]: debug: auto_failback -> 0 (off)
10:53:23 uspstms1 ipfail: [9494]: debug: Setting message filter mode
10:53:23 uspstms1 ipfail: [9494]: debug: Starting node walk
10:53:23 uspstms1 ipfail: [9494]: debug: Cluster node: 56.217.210.50: status: ping
10:53:24 uspstms1 ipfail: [9494]: debug: Cluster node: uspstms2: status: dead
10:53:24 uspstms1 ipfail: [9494]: debug: [They are uspstms2]
10:53:24 uspstms1 ipfail: [9494]: debug: Cluster node: uspstms1: status: active
10:53:24 uspstms1 kernel: send_arp uses obsolete (PF_INET,SOCK_PACKET)
10:53:24 uspstms1 kernel: NET: Registered protocol family 17
10:53:24 uspstms1 ipfail: [9494]: debug: Setting message signal
10:53:24 uspstms1 ipfail: [9494]: debug: Waiting for messages...
10:53:43 uspstms1 ipfail: [9494]: info: Link Status update: Link uspstms2/eth2 now has status up
10:53:43 uspstms1 ipfail: [9494]: info: Status update: Node uspstms2 now has status init
10:53:43 uspstms1 ipfail: [9494]: info: Status update: Node uspstms2 now has status up
10:53:44 uspstms1 ipfail: [9494]: info: Link Status update: Link uspstms2//dev/ttyS0 now has status up
10:53:44 uspstms1 ipfail: [9494]: info: Status update: Node uspstms2 now has status active
10:53:44 uspstms1 ipfail: [9494]: debug: Other side is unstable.
10:53:44 uspstms1 ipfail: [9494]: debug: Other side is now stable.
10:53:44 uspstms1 ipfail: [9494]: debug: Got join message from another ipfail client. (uspstms2)
10:53:44 uspstms1 ipfail: [9494]: debug: Found ping node 56.217.210.50!
10:53:44 uspstms1 ipfail: [9494]: info: Asking other side for ping node count.
10:53:44 uspstms1 ipfail: [9494]: debug: Message [num_ping] sent.
10:53:49 uspstms1 ipfail: [9494]: info: No giveup timer to abort.
10:54:36 uspstms1 ipfail: [9494]: debug: Got asked for num_ping.
10:56:10 uspstms1 heartbeat: base64_to_binary: input length invalid.
10:56:45 uspstms1 last message repeated 6 times
10:57:50 uspstms1 last message repeated 17 times
10:58:46 uspstms1 last message repeated 87 times
10:59:33 uspstms1 heartbeat: base64_to_binary: input length invalid.
10:59:36 uspstms1 last message repeated 98 times

ha.cf contents:

logfile	/var/log/ha-log
debugfile /var/log/ha-debug
udpport	694
baud	9600
keepalive 10
deadtime 30
deadping 30
warntime 20
initdead 60
serial /dev/ttyS0
ucast eth2 10.0.0.1
ucast eth2 10.0.0.2
auto_failback off
node uspstms1
node uspstms2
ping 56.217.210.50
respawn hacluster /usr/lib/heartbeat/ipfail
compression bz2
compression_threshold 1
coredumps false
traditional_compression false
realtime off