home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.questions
- Path: sparky!uunet!cis.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!ukma!kherron
- From: kherron@ms.uky.edu (Kenneth Herron)
- Subject: Re: at(1) command fails
- References: <1992Sep10.161250.13257@cbnewsm.cb.att.com>
- Message-ID: <1992Sep10.152538.3908@ms.uky.edu>
- Distribution: usa
- Date: Thu, 10 Sep 1992 19:25:38 GMT
- Organization: University Of Kentucky, Dept. of Math Sciences
- Lines: 27
-
- sfs@cbnewsm.cb.att.com (salvatore.savastano) writes:
-
- >I use the at(1) command to schedule up to about 1000 jobs at 1 minute intervals
-
- >occasionally at(1) starts spitting out an extra message
- >with each job scheduled and continues to do so until the end.
- > ...
- > at: error in message send
- > job 716077800.a at Wed Sep 9 16:30:00 1992
- > at: error in message send
- > job 716077860.a at Wed Sep 9 16:31:00 1992
- > at: error in message send
-
- At and crontab use a FIFO to communicate with the cron daemon. Each
- message is 26 bytes or more, depending on how a struct gets padded.
- The "error in message send" message is printed if the write() into the
- FIFO fails, suggesting that you're submitting enough jobs at once to
- fill up the FIFO.
-
- The cheap solution is not to submit so many jobs at once. However,
- my guess is there's a simpler way to launch these processes without
- queuing up so many at jobs.
- --
- Kenneth Herron kherron@ms.uky.edu
- University of Kentucky +1 606 257 2975
- Department of Mathematics "Thanks to the magic of the slow-motion
- camera, we can now show you how ugly the inside of someone's mouth can be."
-