Portal suddenly down in the middle of publishing tiling service

[yockee]

I am publishing tile map service, with 4 level of detils. Suddenly, Portal website doesn't load ("site cant be reached" error). When I ping to the site, it replies. 

The space of HDD is 500Gb. And there are still lots of HDD space left. The server has 2 physical cores (4 logical cores) and 16GB of RAM. The server host Arcgis Server and Portal.

Why the Portal failed to load ? What do I need to check ? Is it because of publishing tile service ?

thanks

[George_Thompson]

So are portal and hosting server on the same machine?

What is the CPU / RAM of the machine look like during this issue?

[yockee]

Yes, Server and Portal are in the same machine. I just tried to create the tiling again with 10 LODs, starting from City level to more detail levels. The estimated tiles created are only 3 GB. The server crashed again. Te CPU hit 98.5 mark.. Seems I encounter capacity issue. Any workaround on this apart from increasing CPU-RAM (we are running on metered server, AWS) ?  

[CodyPatterson]

Hey @yockee 

Depending on the load and your answer to George_Thompson, you may need to upgrade your environments resources to better support a larger load of processing. It may have rushed all its resources over to the publishing.

I would check the ArcGIS Server logs and Portal logs to verify that nothing had happened on the Warning or Severe level during that time. I've seen the Portal become unresponsive during intensive loads, or the service itself crashing. 

Also, what's the version of Enterprise?

Cody

[yockee]

These are the errors I have minutes before crashing :

The portal has been initialized and configured but is not accessible. The internal portal database does not appear to be running or accepting connections. Restart the portal machine or machines and if the problem persists, contact Esri technical support (U.S.) or your distributor (customers outside the U.S.). at 2025-08-05T09:55:57

Failed to delete item 'https://myserver/server/rest/services/Hosted/building_tile/MapServer' (id:8e1d4240a006478597cbe472abc6614f). Error while deleteing '8e1d4240a006478597cbe472abc6614f' {"success":false,"error":{"message":"Delete Service failed for url 'https://myserver/server/rest/services/Hosted/building_tile/MapServer (Delete service error: Unable to delete now. Service is being cached.)'"}} at 2025-08-05T09:27:39

Failed to delete item 'https://myserver/server/rest/services/Hosted/building_tile/MapServer' (id:8e1d4240a006478597cbe472abc6614f). Error while deleteing '8e1d4240a006478597cbe472abc6614f' {"success":false,"error":{"message":"Delete Service failed for url 'https://myserver/server/rest/services/Hosted/building_tile/MapServer (Delete service error: Unable to delete now. Service is being cached.)'"}} at 2025-08-05T09:27:14

Failed to publish item 'https://myserver/server/rest/services/Hosted/building_tile/MapServer' (id:8e1d4240a006478597cbe472abc6614f). Job not found.. Job ID: 'j527e4a67dd66461ea6cec64d2ad4a121' at 2025-08-05T09:27:08

Failed to get job status for 'https://myserver/server/rest/services/System/PublishingTools/GPServer/Publish%20Portal%20Service/job...'. Job not found. at 2025-08-05T09:27:08

After restarting the server, now it is back online again.

I running Enterprise 11.0 with postgresql as database.

[CodyPatterson]

Hey @yockee 

Looking at your other comments, I believe this may be a resource capacity issue, especially since it went up to almost 100% on your environment. Since you're not currently able to upgrade these resources, you may reach out to ESRI Support to see if they're able to find anything unusual within your environment before it comes to upgrading, but I suspect they may recommend upgrading the resources on your machine, my relatively low use environment (200 users) uses about all of the 12 cores we have given it.

Cody

[MichaelVolz]

Can you increase to 32GB RAM and try again?

[yockee]

Unfortunately No, without increasing the budget. Its running on AWS.. LoL