Adjust to what? Losing a server this week and finding there's a problem? I've read posts by Jim Robinson and John and there is a problem they're having difficulty tracing after a server went down.
Post 21 Thursday, January 21, 2010 10:02:19 PM by John Robinson
Jim Rob 1 and 4 Thu Jan 28, 2010 12:36 pm
Transparency? Not trying to hide anything. The system is down. We don't know what brought it down. If we did, it wouldn't be down. What ever caused our system to crash a week or so ago apparently has crashed it again. First one server gets swamped and goes down, we route around it, then another swamps and goes down, we route around it, etc, etc. We clear the log jams and sometimes have to remotely reboot some of the servers and we finally got everything up and running again on all cylinders and all looks great for awhile. Now it appears to have happened again. We didn't find anything physically wrong with the hardware or software last time. Don't know why it crashed. We didn't feel like we should replace the whole shooting match, don't have the financial wherewithal to do that anyway, and we also do not have the financial wherewithal to reproduce the whole dmn system somewhere else as a standby. Don't know if that would solve the problem anyway. We've had systems problems many times over the last thirteen years and have always managed to eventually isolate the buggy components and overcome the problems. Also over the years our system has grown from running on one single server to running on a pair, then four and eventually requiring nearly a dozen. A couple of years ago we purchased three new servers each having twice or more the cpu power and capacity of some of the older ones, so we took several of the old ones out of service. When you have a large complex multi-processor system handling millions of transactions per day from hundreds of thousands of simultaneous users over the internet you sometimes have problems. We do not have the resources of a large corporate entity. We have only one systems guy. One. John. He takes care of all the hardware, software, programming, maintenance, networking, contracting with services, suppliers, internet filings, everything. He's on call 24 hours per day, 7 days per week, 365 days per year and does one hell of a great job! I have complete faith and confidence in his abilities.
By the way, he also has to fight off the DOS attack kiddies that are always wanting to screw with us. And he does a helluva great job there too!
I’m beginning to suspect frequency interference from a certain Kinko’s in Waco, Texas. Not sure, but I’m just sayin’...
“Adjust to what? Losing a server this week and finding there’s a problem? I’ve read posts by Jim Robinson and John and there is a problem they’re having difficulty tracing after a server went down.”
Looks like you’ve been here since 2000 (as have I, and I was lurking long before that). Having a server problem “this week” is hardly the issue. Jim Rob and Co. do a fantastic job under the circumstances. It’s just unfortunate that it appears that (based on experience to date) the more successful the site becomes the more unreliable it will be.