This error appears to be an out of memory error. The server had 16GB ram, but we bumped it up to 32GB ram. Also increased the number of cores from 4 to 8. I believe that this is within the 32bit environment. What else can I do to help prevent this from occurring?
DEBUG: 02:02:55.207 [0002] Starting plugin PDF Splitter - 2:02:55 AM
ERROR: 02:02:55.207 [0002] W3001 : Error while executing plugin: Not enough memory resources are available to process this command
ERROR: 02:02:55.207 [0002] PDF Splitter: W1603 : Plugin failed - 2:02:55 AM (elapsed time: 00:00:00.000)
This server has a few OL Connect resources, but works with PDF files as input data. There are a bunch of PowerShell scripts that connect/upload to Sharepoint 365 from Workflow.
Out-of-memory errors are usually due to one or more of these three reasons.
The amount of data being handled by the Workflow process is too big to fit in memory, bearing in mind the 32-bit memory limitation of about 1.8 GB total for the Workflow service.
The amount of data being handled simultaneously by all Workflow processes is too big. In this scenario, the culprit is not the process where the error occurs but rather is the victim of another memory-hungry process.
A memory leak progressively reduces the amount of available memory, ultimately causing the error whose remedy is to restart the service.
Perhaps counter-intuitively, the third cause is actually the easiest to circumvent, simply by scheduling a regular restart. The other two causes are architectural limitations, making them much harder to solve.
(Yes, having a 64-bit Workflow would go a long way to solve this. But there are quite a few technical and commercial reasons why this hasn’t happened yet.)
For the problem at hand, would you share a little bit more information about the context of the error? E.g. what’s this size of the PDF, its source, how the PDF splitter is configure, the size of the corresponding metadata file if applicable, what else is runnng at the same time, etc.