Version

Argent Advanced Technology 3.1A-1401-E or below

Date

Wednesday, 26 Mar 2014

Summary

When an Argent Console Motor is offline, in some rare combinations, it can cause other running Motors fail to fire alerts

When inspecting the Argent Console service log, customer can see lines similar to following repeated over and over in the service log:

23 Mar 2014 06:40:51.208 ISMON1 CORP\admin_sched Potential DeadLock Detected



Thread (Id=0X1870) has waited for more than 310 seconds trying to lock the resource 



                        Lock Name: AAC_IncJobNumber 



                        Acquiring: AAC_ProcessAlert.cpp#127 



                        Used by: AAC_ProcessAlert.cpp#127  

Technical Background

If the offline Motor shuts down unexpectedly such as in case of power outage, and the Motor happens to be holding the cluster lock, the lock will not be released until the Motor is back online

In such a rare situation, the running Motor will have long delay each time to check the lock status

Combined with slow network connection and SQL operation during the unusual period, the total time to acquire lock can exceed the deadlock detection threshold (300 seconds by default), the running Argent Console will recycle and fail to fire alerts

The issue is addressed in Argent AT 3.1A-1401-T5

Resolution

Upgrade to Argent Advanced Technology 3.1A-1401-T5 or later