LLT delayed & lost syslog.log msgs

[Michael S. Stuart]

I just received some new HP-UX 11.11 U 9000/800 9000/800/rp7420 systems configured
w/ ORACLE RAC/VCS for which I am getting repeating syslog.log messages like the following:

Jun 15 13:45:19 fhp025b vmunix: LLT:10019: delayed hb 1000 ticks from 1 link 1 (
lan1)
Jun 15 13:45:19 fhp025b vmunix: LLT:10023: lost 19 hb seq 531079 from 1 link 1 (
lan1)

AND also get dmesg response full of similar messages
LLT:10019: delayed hb 3500 ticks from 1 link 1 (lan1)
LLT:10023: lost 69 hb seq 531329 from 1 link 1 (lan1)

Anyone know what these are/what they mean?
Since there are so many I'm concerned that there may be a problem brewing.

[Kenan Erdey]

hi,

how is the heartbeat connection between cluster noded ? cross, via switch ? it says heartbeat ticks are delayed or lost.

[Michael S. Stuart]

via switch

[Chauhan Amit]

LLT is used to carry GAB messages as well as heartbeat (to tell the node
in the cluster is alive).

If the machine is very busy, and LLT does not get the opertunity to
heartbeat, you will see those messages. It is also possible that the
link might be too bus

-Amit

[spex]

Michael,

Is there congestion on the switch connecting the nodes?

PCS

[Michael S. Stuart]

found that a switch reconfiguration fixed the problem