We are experiencing random workflow failures with no obvious reason and no detailed error messages (other than "an error occurred in workflow_name"). Last failure was at a FlexiTask action - a task had been created and was waiting for an response, an hour later the workflow fails and the task is set to "no longer required".
Here's some background on the environment:
- the workflow has 65 actions and is 950KB in size
- some of those actions are "Log in History List" and "Commit Pending Changes" which we have added along the way while trying to prevent failures
- workflow starts at item creating and there are only a few items created per day
- currently ShP is used for 2 workflows (and nothing more). It's on a all-in-one box (ShP 2013 Foundation and SQL 2014 Express) with just 8GB RAM.
I know it's easiest to blame it on the hardware and database but I still needed some opinions.
As you mentioned hardware does look like an easy target and from the software you have installed it is foundation + sql express? Is this a production farm that you are experiencing this or is this just a trial version running on dev?
Could you try to enable verbose logging (in workflow settings ) on your workflow and republish it ? and check ULS logs to see if you see anything special when the workflow errors out).
I believe you are also enabling the verbose logging at the workflow as that should give you little more details. If you are enabling SharePoint verbose logging then that also could be useful as it would probably add more details on ULS.
Yes, I did this from CA -> Nintex WF Management -> General Settings
The option you refer to (from Workflow Settings) was already enabled (from previous attempts to fix the issue) and I didn't need to re-publish the workflow.