RE: snif.. uptime.

Posted in 2009

Southwell Schuyler — Fri, 30 Jan 2009 08:10:48 -0700 — source: Usenet: comp.databases.informix

>>From: Ian Michael Gumby >>Subject: RE: snif.. uptime. >> >>Just a quick question about considering uptime... >> >>If you have two machines that are set up for HA, and you're forced to take one down for maintenance, >>do you still count that against its uptime because the other machine is still running? Well, for our MFR statistics (apply your own interpretation to the acronym), we report continual uptime, because our users are not affected. Fail over to the HDR instance during maintenance / upgrades; no one is the wiser. That is also why we have multiple WebSphere servers in a cluster, multiple HTTP servers, a redundant 'load balancer', etc. However, I, too, get a tear in my eye when I have to take Informix down. With IDS, I've never gone over about 500 days (and with the nearly clockwork release cycle from IBM, its more like about 180 days or so lately), but that is FAR more that any other piece of software, WebSphere AppServer especially. I have to restart my WebSphere cluster frequently to keep it running well. Informix IDS? Shoot, it's easy to forget about it at times. About the only other 'set it and forget it' server is my MQ server. Schuyler L. Southwell Software Architect (them's fancy words for DBA) Maricopa County Attorney's Office tel: 602 506 8107 southwel@mcao.maricopa.gov <mailto:southwel@mcao.maricopa.gov>