Select to view content in your preferred language

After upgrading 11.1 arcgis server is not responding

3322
15
05-02-2023 07:30 AM
AvinashDalvi_
New Contributor III

EC2 server where I upgrade its work there ArcGIS server but what I did I copy same machine and created another three machine everywhere facing issue. ArcGIS server is not responding. I tried restart and stop doesn't work. Also reauthorisation with new license 11.1 is completed successfully. 

With same upgrade process ArcGIS portal and Datastore is working fine. Only issue ArcGIS server. 

Error : 

"Hmmm… can't reach this page" 

Locahost refuse to connect

 

When I checked window-> services then ArcGIS server service is getting stopped continuously i tried to start bug again stopping. how to check what is issue ?
Tags (2)
0 Kudos
15 Replies
MichelleCasey2
Esri Contributor

Thanks for raising this query.

The are some important Considerations for upgrades when upgrading to ArcGIS Enterprise 11.1 such as .NET. 

Please see some additional documentation here: Upgrade ArcGIS Enterprise which contains a link to Upgrade ArcGIS Server

Hope this helps.

0 Kudos
MichaelSnook
Occasional Contributor III

Hi Michelle

I currently have a ticket in process about a similar issue and it's been lingering for the last 2 weeks.  I'm posting here to get as much help as I can before finally rolling back to a VM snapshot of 10.9.1.  I 'upgraded' my production enterprise to 11.1 on 5/6.  Since then ArcGIS server crashes/stops responding about 2-3 times a day.  The only thing that seems to 'fix' the issue is rebooting the server.  I have scoured ArcGIS server logs, windows server logs, and nothing seems to inidcate that a crash is coming or what causes it.  There have been no changes to the server itself since the upgrade and I never had this issue when 10.9.1 was running.   Please keep in mind that I have done everything in the documentation for the migration to 11 from 10.9.x.  All SOEs have been upgraded, server hardware is at least min spec etc.   Oddly, the installation/'upgrade' went without a hitch.  Also, Portal and the Datastore all seem to be fine...it's just that Server stops responding altogether.  I can't reach server manager or the rest endpoint either internally or externally.  I'm basically having to watch my system all day to see if it's still running and reboot if it's dead so I can keep my production services up.

Please help.

Thanks.

 

0 Kudos
PDSGIS_Department
New Contributor II

Same thing is happening with our installation.

0 Kudos
MichaelSnook
Occasional Contributor III

Do you have any ideas or suspicions to the cause?  I've been trying to check every possible angle and when I fix something that I think is the culprit it crashes again. 

Thanks!

0 Kudos
PDSGIS_Department
New Contributor II
The server ran for days without issue in a test environment. It was only after we made it the production server that the issues started. One of the suggestions from tech support was about increasing the pagefile:
“ Check Page file size across the machines. Default is system managed which is usually a lot less than best practice. I would consider increasing the page file size to a minimum of 1.5X RAM but close to 3X RAM is best. “
We are testing this currently but still looking for other solutions.
0 Kudos
MichaelSnook
Occasional Contributor III

Testing this out over the weekend and (I don't want to jinx it but) I haven't seen a crash since I upped the pagefile size...it was rather small at 5GB.  I bumped it to 16 and hoping that's a fix!

MichaelSnook
Occasional Contributor III

Unfortunately it crashed again early Monday morning. Waiting for support again...

0 Kudos
PDSGIS_Department
New Contributor II
Same here. Support is stumped. We're going to put our 10.9.1 server back into production.
0 Kudos
MichaelSnook
Occasional Contributor III

Still stumped here too...there is nothing in the logs, no error reports at all to even begin to diagnose.  I upped the page file and still crashes.  I removed SOEs, still crashes.  At least now it's only once a day which leads me to believe theres a memory issue when the traffic spikes but my VM stats -- RAM/CPU/Network-- appear to be normal when the crashes happen...I will have to roll back to 10.9.1.  

0 Kudos