Skip to main content

Hi,

We have a strange issue in our FME Flow 2024.2.3, Build 24825 - win64 that prevents our process running in FME FLow. However, same process runs perfectly normal in my desktop and also in FME Server hosted machine. The job fails with no message on why it failed with below line hanging on for few minutes.

Writer `FeatureWriter_2_0' of type `GEODATABASE_FILE' using group definition keyword `FeatureWriter_2_0_DEF'”

Job status: Failure

 

Things i have done so far with no clue:

  • Process runs normal in any PC and also in Server machine
  • Have checked Esri Licenses and they are all valid (as this runs in desktop when run manually)
  • Service account that runs the process has all access required.
  • Process runs if read file (SQL executor) and gdb writer separately, but they dont work and crashes when run in same workspace.
  • Resources -> Logs -> engine -> current -> fmeprocessmonitorengine.log
    • Wed-20-Aug-2025 07:12:52.155 AM   INFORM   Thread-6405   localhost_Engine1   FME_INSTANCE_NAME localhost_Engine1
      Wed-20-Aug-2025 07:12:52.155 AM   INFORM   Thread-6405   localhost_Engine1   Registering with FME Flow on host 'localhost'...
      Wed-20-Aug-2025 07:12:52.171 AM   INFORM   Thread-6405   localhost_Engine1   Successfully registered with FME Flow on host 'localhost'
      Wed-20-Aug-2025 07:14:47.249 AM   INFORM   Thread-6400   localhost_Engine2   SQLExecutor (QueryFactory): 1 input initiators processed
      Wed-20-Aug-2025 07:14:47.343 AM   INFORM   Thread-6401   localhost_Engine2   Aug 20, 2025 7:14:47 AM com.microsoft.sqlserver.jdbc.TDSTokenHandler onColMetaData
      Wed-20-Aug-2025 07:14:47.343 AM   INFORM   Thread-6401   localhost_Engine2   INFO: ConnectionID:1 ClientConnectionId: 3bce5c48-65a2-4461-8aec-762db60bf0d0: SQLServerConnection.supportsTransactions: Discarding extra metadata which can be a result of SHOWPLAN settings:  TDS_COLMETADATA (0x81)
      Wed-20-Aug-2025 07:14:53.252 AM   WARN     localhost_Engine2   393562 : Process "localhost_Engine2" ended unexpectedly. Being restarted on attempt 1...
      Wed-20-Aug-2025 07:14:53.253 AM   WARN     localhost_Engine2   393563 :
  • Have checked Resources → Logs→ core→ current → fmeserver.log

Wed-20-Aug-2025 07:14:53.037 AM   ERROR    engine-localhost-localhost_Engine2   401844 : Job 59932: Could not read FME Engine response. Connection may have been lost. Please see: http://fme.ly/e100005#could-not-read-engine-response
Wed-20-Aug-2025 07:14:53.037 AM   ERROR    engine-localhost-localhost_Engine2   401872 : Job 59932: Failed to get translation result. Returning failed result.
Wed-20-Aug-2025 07:14:53.071 AM   WARN     engine-localhost-localhost_Engine2   402682 : Job 59932: Failed and has reached the retry limit of 0. It will not be resubmitted again and a failure result is being returned.
Wed-20-Aug-2025 07:14:53.076 AM   INFORM   engine-localhost-localhost_Engine2   401936 : Disconnected FME Engine: localhost_Engine2 on host localhost
Wed-20-Aug-2025 07:14:53.641 AM   INFORM   requesthandler   401831 : Accepted new engine connection from /127.0.0.1 on port 7070
Wed-20-Aug-2025 07:14:53.733 AM   INFORM   fme-engine-registration-1   401937 : Accepted new FME Engine connection.
Wed-20-Aug-2025 07:14:53.733 AM   INFORM   fme-engine-registration-1   401950 : Registering FME Engine...
Wed-20-Aug-2025 07:14:53.740 AM   INFORM   fme-engine-registration-1   401842 : Transaction port 52695 opened for FME Engine communication
Wed-20-Aug-2025 07:14:53.751 AM   INFORM   fme-engine-registration-1   401983 : Transaction channel socket keep alive: true
Wed-20-Aug-2025 07:14:53.754 AM   INFORM   fme-engine-registration-1   401864 : FME Flow ready to process requests using instance localhost_Engine2 

 

Please let me know if any of you have similar issue, or there is bug that i am not aware of?

Subodh Dhakal

 

The logs show that the engine crashes, hence the restarting.

A couple of extra things i’d also check:

  • Are there any errors in Windows Event Viewer?
  • Can you replicate the issues if you setup another workspace that just reads and writes to/from those locations? Set this up from a blank workspace, don’t copy and paste
  • Have you tried completely removing the readers/writers (not just the feature types) and adding them in fresh?

I’d raise it with Safe regardless, as you suggested, it might be a bug

 


Thank you ​@hkingsbury 

That’s right, engine is crashing with no apparent message to troubleshoot. No errors in Windows event viewer, I have tested your second point and third point too with no success.

I raised with Safe and safe mentioned that this is known issue, see Safe message for workaround below: 

This sounds like a known issue we have with using Esri Geodatabase with another JDBC connection. When using JDBC drivers and SDE Geodatabase in the same FME workspace on FME Flow, particularly on a system where ArcGIS Server is also installed, there can be a conflict with Java Virtual Machines (JVMs).
 
This can result in either:

  • JDBC drivers failing to work, or
  • The workspace crashing when accessing the SDE Geodatabase.

 

Workarounds to consider:

  1. Use a non-JDBC format (if available):
    If the format you're using via JDBC (e.g., SQL Server) is also supported natively in FME without JDBC, switching to the native format can avoid this conflict.
  2. Split the workspace:
    If no non-JDBC alternative exists and both JDBC and SDE Geodatabase must be used, consider separating them into two workspaces. These can be chained together using Automations or a parent workspace with FMEFlowJobSubmitter transformers to call the individual components.
  3. Stop ArcGIS Server services (best but not always practical):
    If ArcGIS Server is installed but not actively required, stopping its services will prevent the JVM conflict. FME Engines only require ArcGIS Server to be installed (and licensed), not actively running. Disabling ArcGIS Server removes the JVM clash and allows both JDBC and SDE components to function correctly.

Reply