On Wed, May 23, 2012 at 6:18 PM, Stephan Beal <sgb...@googlemail.com> wrote:
> scripts to be used. IF so... you could cron a job (outside of jenkins) > which > OTOH... (kind of off-topic but it's related...) a few years back i was administering Solaris systems at a bank. Those systems ran various stock-market-processing software, mostly transferring trades between brokers. On one system i noticed that one of the monitoring jobs, a simple "ping"-based script provided by the app developers (i was the sysadmin), had about 40 stale instances running, some of them months old. So, like any good admin, i fished out their PIDs and killed them. Two minutes later the app's developers called me in a panic because their server software had spontaneously restarted and switched cluster nodes. What had happened was quite interesting, from a sysadmin's/software developer's standpoint (and we're quickly approaching the point of this whole side-note)... cron, at least on Solaris, has a limit of how many jobs it can run at once and queues up jobs if the limit is reached. When i killed the 40-some-odd ping processes, 40-some-odd _other_ queued processes, some of them several weeks old, were finally run. That included, for example, the nightly shutdowns and morning startups of the app servers (those ran several times in a row, and i couldn't do a thing to stop it). The application's project manager was quite upset. My point is simple... cron can also have its drawbacks. -- ----- stephan beal http://wanderinghorse.net/home/stephan/ http://gplus.to/sgbeal
_______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users