I have had a suspicion that the external action can lock against itself under load. Years ago I was using a multi instance setup and found that the servers locked up in a way that you are describing. I went to a virtualized configuration with one instance per virtual and the problem went away. This was on TS 6. I also had a Witango 5.5 user complain of something similar. In that case, adding servers solved the issue.
Other than those experiences and a hunch about the root cause being related to the external action, I don’t have much more to go on. One thing that I’ve done which I think helps is to write out a new batch file for each external call. This creates two situations: 1- Since I swap all the variable data in the file write, I don’t use environmental variables 2- Since I’m writing a uniquely named file, no two external calls access the same file Furthering my hunch, I think the problem is tied into the environmental variables. Since TS8 is completely rewritten in Java, and uses the Java Process system it won’t have any problems. Robert From: Peter Dobbs [mailto:pe...@enginess.io] Sent: Thursday, July 12, 2018 8:31 AM To: TeraScript-Talk@terascript.com Subject: TeraScript-Talk: external action failing Have you ever experienced a situation where external actions (cmd line windows) threw errors for no reason? We have code that has been stable for many years. Recently we’ve begun to dramatically increase the number of cli interactions and it seems as though windows stops accepting requests from the app server. Peter _____ To unsubscribe from this list, please send an email to lists...@terascript.com <mailto:lists...@terascript.com> with "unsubscribe terascript-talk" in the body. ---------------------------------------- To unsubscribe from this list, please send an email to lists...@terascript.com with "unsubscribe terascript-talk" in the body.
