MIDServer often goes down
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎10-03-2017 06:03 AM
Hello guys,
I'm experiencing some anomalies on the servicenow instance.
My unique MIDServer often goes down. Which could be various causes?
The computer that hosting the midserver is always on...
Thanks.
Regards,
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎10-03-2017 06:17 AM
Hi De Mitri,
Following links should help you debugging:
ServiceNow KB: Troubleshooting a MID Server that is down or not responding (KB0535180)
ServiceNow KB: TROUBLESHOOT: MID Server upgrade issues (KB0596459)
Mark Correct if this solves your issue and also hit Like and Helpful if you find my response worthy based on the impact.
Thanks
Ankur
Ankur
✨ Certified Technical Architect || ✨ 9x ServiceNow MVP || ✨ ServiceNow Community Leader
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎10-03-2017 06:25 AM
Hello Gaetano.
The logs of the MID server should tell you more, sometimes directly the cause.
Here are links to documentation on various scenarios:
Troubleshooting a Down ServiceNow MID Server
ServiceNow KB: Troubleshooting a MID Server that is down or not responding (KB0535180)
ServiceNow KB: TROUBLESHOOTING GUIDE: MID Server (KB0597571)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎10-03-2017 06:29 AM
To add to this, create a business rule on the MID server table that opens an incident when the status changes to down. That will help you in your troubleshooting efforts, I think, to have an approximate time. Often, when troubleshooting midserver crashes, it helps to be able to go directly to the time in the logs, as shown by the time the incident is created.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎10-03-2017 11:57 AM
Mid servers go down when the instance can't get a heartbeat back.
Every 5 minutes, a HeartbeatProbe is sent to each mid server. When the ecc_queue input response comes back, the mid is marked Up. If no response has been received since the last heartbeat was sent 5 minutes ago, the mid is marked Down.
So, you can look at the ecc_queue for your mid server and see how responsive this mid has been to HeartbeatProbe's, identify when it has been having problems, and then look in the agent/logs directory at the entries in the agent0.log.0 and wrapper.log files for that time period to get an idea what's going on.