Just beware!
I spent the whole day debugging a process with a lot of stuff happening inside, going out of memory.
At the and of the day, the “solution” was to set the Workflow logging level from “Task success and failure with details” to “Task failure”. After that everything behaved as expected!
Argh!
-i.
May I suggest that if your memory problem were due to the logging level of Workflow, you might need to increase you memory?
The logging only occurs once the process is finished. It is not written in live time. And for this to make any difference, you would need a production environment that has a very high load, hence, increasing the memory resources.
There is no way to increase the memory for the 32bit workflow process.
As far as I remember (my initial post is almost 10 months old), the problem was in the excessive logging of the database query tasks with custom SQL.
The logging occurs all the time (in memory?) and is only flushed to disk, when the process is finished…
So as said, if one cannot increase the memory, one can spare some by reducing the workflow logging 
Oh! so you had memory issue with the Workflow itself? Sorry, thought you had memory issue with the whole server…my bad…
Hi,
Which section of workflow did he set this? Is this from server configuration? As cant find this settings in workflow preference.
Thank you
Upper left corner big W->Preferences->Plu-in->General