Select to view content in your preferred language

Widespread ArcGIS Online hosted feature layer failures (publish, offline sync/replica, and export/backup)

182
0
Wednesday
Spengler
Occasional Contributor

We are experiencing organization/account-level failures in ArcGIS Online (user: cemml_alexs, Colorado State University) affecting core hosted feature layer workflows since approximately August 5, 2026, with major failures observed on August 10–11, 2026. Specifically...

Publishing and exporting hosted feature layers fails (ArcGIS Pro and AGOL browser), including minimal test data and simple schemas.

Pro logs show staging succeeds, SD upload starts, then publish fails with generic failed status. Often it will just stall out at "uploading service definition" in ArcGIS Pro.  

Screenshot 2026-08-12 065236.png

Attempting to publish to AGOL through a browser also stalls out.

image (15).png

Often the service definition and feature layer will be visible in ArcGIS Online but return a generic error message when viewed.Screenshot 2026-08-11 161516.pngimage (16).png

Rarely, attempting to publish will actually fail in Pro and return an error message like thisimage (14).png

After failing, the feature layer or service definition is visible on ArcGIS Online but will return “item not found,” when opened.

Additionally, any Export to file Geodatabase in AGOL or Export Features in Pro will not complete after hours of sitting at "Exporting" in AGOL or "100%" in Pro. 

Spengler_0-1786575790529.png

 

Offline area creation and sync fail across multiple hosted layers (ArcGIS Pro + Field Maps).

Errors include: internal replica errors, inability to create/sync replica, HTTP 500, ERROR 004005, ERROR 004547. These errors are more common in our org but usually they can be resolved.Screenshot 2026-08-12 065919.png

Screenshot 2026-08-12 064357.png

This coincides with an error that we've never encountered before related to "The database ʻdb_660ʻ has reached its size quota. Partition or delete data, drop indexes, or consult the documentation for possible resolutions. 

Spengler_0-1786555749677.png

 

image (17).png

image (18).png

We've now encountered this on multiple projects over the last two days and have successfully removed enough replicas to allow edits and clear this issue BUT we still cannot manage to taking those maps offline.

Nightly backup/export automation is failing at scale.

For several years we've run a daily script which exports hosted feature services to FGDB, downloads, then deletes export items. Now it's repeatedly failing with missing exportItemId and timeouts; example:
2026-08-10 23:11:56,756 ERROR Attempt 3 failed for item 'TA21RodentControl'. Error: 'exportItemId' Screenshot 2026-08-12 070442.png

What we already validated

User role/privileges include hosted publishing/admin rights. My User Type is
Professional Plus and my Role is CEMML-HI Admin which has publishing rights. Other accounts with same User Type and Role also experience these same issues.

Network issues. I've tried publishing outside of my network and encounter the same issues.

Fresh GDBs, tiny datasets, unique names, and browser-based publish attempts still fail.

Sync prerequisites are in place (GlobalIDs, attachments, related records where needed).

Any advice on how to proceed with any of these issues?

Best,
Alex

 

0 Kudos
0 Replies