[Linux-HA] HA2 does not always increment fail-counts

[Oliver.Jansen at t-systems.com]

Hi,

I'm running HA-2.0.6 compiled on a SLES9 and encountered a possible bug.

Normally when resources fail and heartbeat finds out with the monitor action it increases the fail-count for that resource on that particular node.

But when failures occur by turns on node A and node B, the fail-counts are not increased. I was able to reproduce this several times and tracked it down to the tengine where it tries to find matching action events but cannot find the appropriate action. Unfortunately I did not quite understand the code and its context and neither have much time to dig deeper into it.

Has anybody else encountered this and/or is this maybe fixed already? 

Nevertheless I will try to reproduce this with heartbeat-2.0.7 and provide some more information.


Thanks in advance,
Oliver.

-- 
Oliver Jansen
T-Systems
Systems Integration
System Engineer

Project Center:            Air & Travel
Solution & Service Center: Services
Industrie Business Unit:   Services & Finance

T-Systems GEI GmbH
Address:        Pascalstraße 8,   52076 Aachen
Postal address: Postfach 500 144, 52085 Aachen

Phone:     +49 (2408) 943 1926
Faxsimile: +49 (2408) 943 1430 
E-Mail:    Oliver.Jansen at t-systems.com
Internet:     http://www.t-systems.com