From: Joe Z. <joe...@za...> - 2004-12-06 17:35:09
|
Jochen Metzger wrote: >Hi Rene, >Hi Joe, > >i am inplementing for a customer and I thinking about logging stuff what >we talked about a couple of months ago. > >Mainly I want: > >1. Have logging what has be updated and what failed. >2. Be sure that bobs is still running properly > >as I work with companies and then let things running >and maybe get notified when something went wrong. > >I have a couple of hours of budget to do some stuff >and I want to discuss it with you first to then put >it in the cvs afterwards. > >At present I wrote a workaround (quite a quickhack) >that writes a timestamp and checks that. >(I have to figure out, how I did that and if it makes >sense to the main version) > >Any ideas ? > >Jochen Metzger > > > Have you got the latest version from cvs? It writes all output to bobs.log and includes the timestamp. I get the output from cron every day and I also run this cron command to give me a brief(er) listing: head -n 1 /var/log/bobs.log; /bin/grep -E ': Starting backup|: wrote |: total size is ' /var/log/bobs.log; tail -n 1 /var/log/bobs.log It's helpful, but I was thinking of adding real error checking to the backup scripts to say things like "successful backup of <server/share>" or "backup failed for <server/share>", with optional email notification. But I won't be getting to it anytime soon. To know if bobs is not running would require a 2nd cron job to check something, like the timestamp you write. I'm thinking that's what you have in mind. That's not a bad idea for monitoring your customers system. Joe |