Windows service with NService Bus SQL Server Transport stop dequeuing message

, ,
1

I have windows service where i using NServiceBus. Randomicaly NServiceBus stop dequeuing message from queues and for restart dequeing i need to restart the service.

On NLog config i have setup the info log level and i can’t find significatice log.

I use:

NServiceBus version: 7.5.0

NServiceBus.Persistence.Sql 6.2.0

NServiceBus.Transport.SqlServer 6.3.1

I can not activate trace log for transport because the daily log file was etremely big

2

Hi,

Does this only happen in the one application or does it happen in more than one?

Based on the transport and persistence, there might be some clues in the log files related to SqlConnection failures.

If the log file is set to default log levels, are there any WARNs or FAILs in the log file? Anything to give us a clue to get started.

3

The log levels is set to default but in the file there aren’t anything, if you have suggest for log something is perfect. The problem happen in the one application

4

Do the other applications connect to the same instance of SQL? Can you look at SQL and see what applications are executing what scripts when it stops?

sp_who and sp_who2 would be helpful here.

5

I checked the other instances the scripts they run and none of those impact the tables of the service that crashes

6

Can you look at sp_who and sp_who2 when the application is stopped to see the server is doing?

7

I’m experiencing the same behavior — the endpoint silently stops dequeuing messages with no errors or warnings in the logs. The only resolution is restarting the service, after which processing resumes normally. In one instance, over a million messages had accumulated in the queue and it took three restarts of the host to fully clear the backlog.

Unfortunately, I wasn’t able to run sp_who or sp_who2 at the time of the incident, but I’ll make sure to capture that information the next time it occurs.

Is there any additional logging or diagnostics that can be enabled on the transport level to help identify the root cause? Any configuration that might surface what’s happening when the dequeuing stalls would be greatly appreciated.