From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <2f1b9639cec7e8fd4225874b7bd81b5d@quintile.net> From: "Steve Simon" Date: Sat, 3 Jun 2006 12:18:50 +0100 To: 9fans@cse.psu.edu Subject: Re: [9fans] *poof* In-Reply-To: <20060602222546.91d80400.20h@r-36.net> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Topicbox-Message-UUID: 5a30dd54-ead1-11e9-9d60-3106f5b1d025 > I had the same problem with ddr.9grid.de. It ran out of memory short > after cron was started. For now cron is turned off there, but that is > no real solution. I too had problems with cron, It seemed that timesync can set the current time back a long way (days) in some circumstances and when it discovers its error and corrects the system clock cron goes nuts trying to run all the jobs that have been missed. This only happened once a month, and somtimes at system boot so it was difficult to debug. I looked at timesync but it looked correct to me. I submitted a patch to cron which should have prevented the madness (only 24 hours of catchup performed at most), and though it helped, it didn't seem to cover all cases. I have now I delayed the start of cron until timesync has (hopefully) stabalised the system clock, which seems to solved the problem. felix% grep cron /rc/bin/cpurc { sleep 600; auth/cron >> /sys/log/cron } & not really a solution, more of a workaround. -Steve