Hi,
I'm experiencing an issue with SystemCaller when publishing to FME Server (2019.2 - build 19801 - 64-bit))
We're using a SystemCaller to execute a program (with some parameters supplied) which is stored on a shared network, and it works successfully on FME Desktop 2019.2 (same build number as the Server instance). The Server has access (read/write) to the same network location, but when the published workspace runs, it just stops at the SystemCaller Execution. No error, no falling over...it just hangs. We left it overnight, but it was still the same the next morning.
Things we've tried, to narrow down the issue:
- Ran the script in SystemCaller on FME Desktop - it worked in around 30 seconds
- Ran the same script in the cmd prompt on the Server box (as svc_fme user) - it ran successfully in 30 seconds
- Tried hardcoding the network location into Server (rather than using $(FME_SHAREDRESOURCE_DATA)) - this didn't affect the outcome at all - still hung
- There are additional '/'s in the UNC path (as supplied in the $(FME_SHAREDRESOURCE_DATA)) - we tried removing these but it made no difference
For reference, the script we're trying to run is (with the password hashed out):
//corp/Landmark/Groups/Data/FME/FME_Server_2019.2///resources/data/Putty/plink.exe -ssh genasys@exconv02 -pw ###### cp -R /storage/50kcrm-35/MERGED /imogen_raster/50kcrm/l00035 > //corp/Landmark/Groups/Data/FME/FME_Server_2019.2///resources/data/Putty/Output/OutputDEL2.txt
We think it might be something to do with running cmd prompts in SystemCaller using a non-Admin account - any ideas how we force FME Server to run it as Admin?
Or has anyone any other ideas we can try?
Thanks,
Martyn