We've had some cases in the agent code where the agent
doesn't properly release itself back to the broker
when issues arise, leaving the watchdog to follow up
with pings after the fact. However, once I added some
tracing logic, it seems like the watchdog is not
always cleaning up properly. I've added some "RUN
pReleaseAgent." statements in the message check and
password check logic in gnm_agt.p. I intend to update
this bug as I gather more information.
Marian Edu
broker
linux
Public
|
Date: 2006-12-08 08:32
|
|
Date: 2006-11-03 22:51 Logged In: YES |
| Field | Old Value | Date | By |
|---|---|---|---|
| status_id | Open | 2007-06-21 13:19 | medu |
| resolution_id | None | 2007-06-21 13:19 | medu |
| close_date | - | 2007-06-21 13:19 | medu |