Self hosted upgrade database error

petterbj_rgmo
Tera Contributor

‎02-22-2018 11:12 PM

Hello everyone,

We are trying to upgrade our self-hosted instance from Helsinki patch 3 to Jakarta patch 7. When upgrading we followed the documentation as provided on ServiceNow HI. Still, when the upgrade is trying to execute, we are getting some errors in regards to "JRobin database". The errors are as follows:

  • 2018-02-22 21:17:24 (658) glide.internal.monitor SYSTEM WARNING *** WARNING *** JRobin database: sql_stats does not exist : Could not open sql_stats.89733333b51013003eaaecb695694e40 [non existent]
  • 2018-02-22 21:17:24 (660) glide.internal.monitor SYSTEM WARNING *** WARNING *** Unable to create JRobin database: sql_stats.89733333b51013003eaaecb695694e40 : jrobin_definition does not exist: sql_stats
  • 2018-02-22 21:17:24 (669) glide.internal.monitor.writer glide.internal.monitor.writer WARNING *** WARNING *** JRobin database: transaction_response_time does not exist : Could not open transaction_response_time.40343b33411013003eaa60b8c9a089e8 [non existent]
  • 2018-02-22 21:17:24 (670) glide.internal.monitor.writer glide.internal.monitor.writer WARNING *** WARNING *** Unable to create JRobin database: transaction_response_time.40343b33411013003eaa60b8c9a089e8 : jrobin_definition does not exist: transaction_response_time

Snapshot from the system logs:

 

find_real_file.png

This repeats it self over and over again, not allowing the upgrade to move forward. Can anyone explain what JRobin database is in a ServiceNow context? And has anyone any pointers to what can be done to avoid the errors above? As of now, the nodes are stuck trying to upgrade and spitting out these errors all the time. We have also created a ticket at HI, but we are waiting for a technical resource to be available.

 

Preview file
67 KB
0 Helpfuls
  • 1,932 Views
3 REPLIES 3

Patrick DeCarl1
ServiceNow Employee
ServiceNow Employee

‎02-23-2018 07:27 AM

ITs been some time since I had to work on a self hosted instance. When you ran through the upgrade process, you took down all nodes but one and only run the upgrade on a single node until it completed right? Once that node completed, you would go and do the same thing to other nodes for your evn. Also, was there a new GEN process/setup you needed to do from H and J release? 

 

Have you taken the time to open a HI ticket with SN support? 

0 Helpfuls

petterbj_rgmo
Tera Contributor
In response to Patrick DeCarl1

‎02-26-2018 04:40 AM

Hi Patrick,

We took down all the nodes, and we ran the upgrade on one single node. Once we started to run the upgrade itself (don't remember the linux command in my head right now), we started getting the errors above. We let everything run for about 2 hours, we kept getting the same errors and there was no logs in the upgrade history, not the upgrade monitor pointing to any upgrade being run.

We have created a ticket with HI and they are saying that the errors in regards to JRobin_database is about getting performance stats for the MySQL server.

What is strange is that we have plugins which are not available on Helsinki (Automated Tet Framework e.g) ready for installin gon the instance. But on the upgrade history no inout of an upgrade is to be found. This means that something has been done, and other things have not.

Our application nodes seem to be running fine, it is just hard to say where the upgrade fails..

0 Helpfuls

Patrick DeCarl1
ServiceNow Employee
ServiceNow Employee
In response to petterbj_rgmo

‎02-26-2018 06:33 AM

I believe the plugins are stored on the app node not database so that is probably why you see them in the plugin list.  

 

Did SN Support let you know how to fix issue? Does the DB user have full rights to database? 

 

Did SN update the GEN layout any? 

0 Helpfuls