We have noticed that following our upgrade of Optigold(?) (sometime in February), the nightly automated tasks (some of which telnet places and do things) have gone from running sequentially spaced apart to running all at once (at least from the logs)...is that a setting? For instance these are tied to a number of automated cleanup cancellations: ttypj Wed Mar 2 00:08 - 00:09 (00:01) ttypi Wed Mar 2 00:08 - 00:09 (00:00) ttypa Wed Mar 2 00:08 - 00:08 (00:00) ttyp9 Wed Mar 2 00:08 still logged in ttyp4 Wed Mar 2 00:08 - 00:09 (00:00) ttyp6 Wed Mar 2 00:08 still logged in ttyp5 Wed Mar 2 00:08 still logged in And this was before the upgrade: ttyp1 Thu Feb 3 01:00 - 01:01 (00:00) ttyp1 Thu Feb 3 00:58 - 00:58 (00:00) ttyp1 Thu Feb 3 00:55 - 00:56 (00:00) ttyp1 Thu Feb 3 00:52 still logged in ttyp1 Thu Feb 3 00:50 - 00:51 (00:00) ttyp1 Thu Feb 3 00:47 - 00:48 (00:00) ttyp1 Thu Feb 3 00:45 - 00:46 (00:00) ttyp1 Thu Feb 3 00:42 - 00:43 (00:00) Note that it seems, post upgrade, to be opening a whole bunch of telnet windows all at once (thus the varied ttyp numbers) but reusing one (or ending up with one reused) pre-upgrade. Sam
Are you running new or old style events? If it's a new style event, are you using the option to pause after the script is run?
I am using the new style. I do not have anything set in the pause after (and noticed that before I posted my message but reasoned that even if I pause for 5 minutes after a script is run, the problem is that they all run at once, so that would still happen, they would probably all just run at once and then all pause at once, etc...? ...and also that I did not consciously change my scripts as part of the upgrade). I have tracked down in old email the exact date of our upgrade and sure enough it jives. The last autotasks run on the Friday before the weekend we upgraded were all about two minutes apart and all getting the same telnet slot, and the ones on the following Monday night (when auto-tasks was restarted) all ran at exactly the same time and opened a bunch of seperate telnet sessions all at once...and it has done that every night since. Sam
You would want to use the pause option in the event itself... the mandatory pause went away since most events run in a single thread anyway it was a pointless slowdown to force a pause. But some events (telnet for example) and spun off into it's own process. So you would want to use the pause option.
If a telnet is tied to a timed cancellation of an account (expires on date) which scheduled task item determines the time of day that it runs? (I have a "cancel customers every 60 minutes", but conspicuously these run around 12:45am making me think they tie to index building.........unless they are really running every 60 minutes but being based on the date - only the first one really counts, and that perhaps happens after the 12:00am stuff has finished for the first (but really only) time...just wondering if I could set these to go off at 8:00am and watch them, rather than stay up to watch.) Was the expiration of the mandatory pause between events sometime between 3.1.3 and 3.3.6? Is the "Pause After Telnet Trigger" 0:00:00 in "hours minutes seconds" or "minutes, seconds, 100th's of seconds"? Sam
The automated tasks are not threaded... so if you have an event that has a time that collides with another, they are queued to run. The format is hh:mm:ss
So I changed my "Pause After Telnet Trigger" to 0:12:00 (to give it an unmistakably wide difference in the logs) and overnight last night look what happened: Last Night: ttyp0 Fri Apr 8 00:44 - shutdown (07:08) ttyp0 Fri Apr 8 00:42 - 00:43 (00:00) ttyp0 Fri Apr 8 00:40 - 00:41 (00:01) ttyp0 Fri Apr 8 00:38 - 00:39 (00:01) ttyp0 Fri Apr 8 00:37 - 00:37 (00:00) (as opposed to) Pre-Telnet-Trigger-Change: ttyp9 Thu Apr 7 11:09 - 11:10 (00:00) ttyp6 Thu Apr 7 00:56 - 00:57 (00:00) ttyp3 Thu Apr 7 00:56 - 00:57 (00:00) ttyp8 Thu Apr 7 00:56 - 00:57 (00:01) ttypa Thu Apr 7 00:56 - 00:57 (00:01) ttyp7 Thu Apr 7 00:56 - 00:57 (00:01) ttyp9 Thu Apr 7 00:56 - 00:57 (00:01) (which is not the 12 minutes I set it for...but which is suspiciously similar to the old mandatory pause) Pre-Upgrade-to-3.3.6: ttyp1 Thu Feb 3 01:00 - 01:01 (00:00) ttyp1 Thu Feb 3 00:58 - 00:58 (00:00) ttyp1 Thu Feb 3 00:55 - 00:56 (00:00) ttyp1 Thu Feb 3 00:52 still logged in ttyp1 Thu Feb 3 00:50 - 00:51 (00:00) ttyp1 Thu Feb 3 00:47 - 00:48 (00:00) ttyp1 Thu Feb 3 00:45 - 00:46 (00:00) ttyp1 Thu Feb 3 00:42 - 00:43 (00:00) (which is fine, that's the main thing I was after, it's almost just an academic observation which looks, from my admittedly limited vantage point, like enabling the telnet trigger time turns on the mandatory delay but does not obey the time...probably nothing but it may be worth noting that when I first clicked that it changed from 0:00:00 to just 0 so I entered "3" and clicked so it made 3:00:00 so I then reconstructed that number, with 0:12:00 ...just in case it has to be 'made' by OG to work). Sam