All migration work has completed and We Eatt is now on the new servers.
If you still see the maintenance page, try refreshing and/or reopening your browser.
We’ve started our migration at 8:35 PM CST on September 14, 2009.
Will update here when it’s complete.
The We Eatt site now shows a maintenance page.
15 minutes of downtime is expected later today when we move servers. Check the link above for more details; we’ll post here when we are about to move it.
Woke this morning and found we had some intermittent downtime last night after a bad image processing on a Chef’s avatar.
We haven’t seen much downtime in the last two months, so maybe we were due!
Seemed to go up and down, but the site was generally responsive (or maybe responsive to 50% of the requests). Either way, fixed at 7:02 AM CST.
Between 8:00 PM CST and 8:30 PM CST we received reports of unavailability for www.weeatt.com. You can view the Slicehost posting for more information.
Slicehost is great and was able to isolate the denial of service attack, work with Cogent and Level3 and reduce the downtime. Go Slicehost!
We’re starting the upgrade of our server now (7:30 PM CST). We will update here when it’s completed.
Update: We’re all resized and readt for action. (8:41 PM CST)
Technical Explanation
Our hosting provider, Slicehost, logged that we were thrashing our swap file and pro-actively rebooted the server. Since we had been online for over 70 days, our server did a self-check of the harddrive, and that took over an hour to run.
The reason we were thrashing is that I upped our number of Mongrel instances from 3 to 4. With only 512 Mb of RAM, I was hoping we could get a productivity boost. Looks like I was wrong and we’re back down to 3 instances.
Looks like our provider (SliceHost) is having some issues right now, looking into it.
We’ll update more as I know more.
Update: We Eatt is in the process of being brought back online; our provider forced a reboot when the system was “thrashing” its swap file. Slicehost did have a networking problem, but it appears to be unrelated to this outage.
Update: Back Online at 9:00 PM, will post techincal explanation
Looks like some weirdness on the webserver front, I had to restart both Apache and Mongrel at 8:30 PM, CST, on 10/1/2008.
Just got a notice from pingdom that We Eatt was down, not sure what happened. Will look into; restarted mongrel and all is fine now: 9/26/2008 at 12:25 PM CST.
