[ale] RHL9 crond doesn't read in new crontab entries

[Richard Kolkovich]

Jerry Yu wrote:
> I used 'crontab -e'.  I am surprised too, since I never had to restart
> crond to make new entries effective on Linux or Solaris. On SCO
> openserver I had to run crontab against the new crontab file. But
> that's a different story.
> Actually I also had this on a RHAS 4.1 (CentOS 4.1) lab server which
> was up 190 days only. 'rpm --verify vixie-cron-4.1-20_EL' only
> reported some doc I purged. anoncron is there too.

Very odd...anacron, IIRC, is not a cron daemon but rather something to 
make sure missed jobs get run (i.e.- shut a machine down for new 
hardware and it misses a cron job time).

I have some RHEL/CentOS boxen around here...I will see if I can 
reproduce your issue.

--

Richard Kolkovich

> 
> On 8/21/06, Richard Kolkovich <sarumont at sigil.org> wrote:
>> Jerry Yu wrote:
>>> I wrote a simple bash script then added it to the cron a RHL9 boxen.
>>> Weird enough, it didn't take. Initially I troubleshooted my shell
>>> script, you know, adding verbose output and such. The /var/log/cron
>>> shows other cron jobs ran just fine. After pulling my hair a little,
>>> it dawned on me that the new job didn't show in the /var/log/cron at
>>> all. The fix, of course is to restart crond.
>>>
>>> Hmmm. I inherited the boxen a year ago. It has been up a bit over  500
>>> days now. I never experienced such problem before with Solaris or
>>> Linux. The symptoms repeated on another RHL9 boxen (up about 483
>>> days). Any idea why?
>> Initially, I would guess that it is a different cron daemon.  How are
>> you adding these jobs?  Are you using crontab -e or editing /etc/crontab
>> directly?
>>
>> I have never had to restart vixie-cron when adding a cronjob via either
>> of those methods...
>>
>> --
>>
>> Richard Kolkovich