RE: Hands on (Was: Looks Like The Party Is Over In Menlo (2nd try))

From: William Rice <ricew@operamail.com> > > >===== Original Message From "Obnoxio The Clown" <obnoxio@hotmail.com> >===== > >From: William Rice <ricew@operamail.com> > >> > >>We had over a 100 boxes with informix on them with automated monitoring > >>software. The systems were not real big, and the applications evolved > >>quite > >>slowly. There were some boxes I logged into maybe four times in my > >>two years with the company. > >> > >>I am in no way commenting on whether or not one should use SE, just > >>addressing the question on a hands off instance of IDS. > >> > >>P.S. a lot of time was spent writing the automated monitoring scripts, >so > >>this definitely required full time knowledgable DBA's(wish I had been in > >>that category ....), but they were completely hands off. > > > >So if something went wrong, you didn't have to do *anything*? If you had >an > >AF, everything would just sort itself out? Very impressive. > > >It didn't happen enough to us for me to consider that an extreme exception. Same with most of my SE sites. I mean, 99.9% of the time, they just ran. And you didn't have to write any scripts up front. >I can honestly say in my 2 years there we had three AF files generated on >production systems, two of which were generated on the same evening, none >of those three caused database corruption. We had two instances which >required oncheck repairs during that same time period. So yes, some >looking >into issues was necessary, but for over 100 instances thats pretty darn >reliable. That is pretty reliable, I've worked on IDS sites (eh, Paul?) where the server crashed daily. And not just because of me, either! >The systems we monitored were mostly out in the field, the power >supplies to these boxes weren't the most stable. There weren't many months >where a couple of boxes didnt have hard crashes due to power loss >or hardware failures, but IDS kept on trucking... So yes, I will say that >once everything was set up we pretty much didnt have to do anything except >add space when the alerts came about, change our scripts when we upgraded, >and try to assist applications. Maybe we were just real lucky. Not necessarily. I had plenty of IDS *and* SE sites that never required intervention. I just think that the anti-SE argument is a bit specious. :-) >P.S. I am not counting issues which required DBA intervention due to > ><SNIP> LOL... ________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com