OTB Update – Going Temporarily Offline Later Today
Commenting will turn off at 1pm ET today
Over the past few months, James has blogged about efforts to make OTB more cost-neutral. Over most of June and July, working around my appendectomy, he and I have made progress on a number of background changes that should greatly help with that. The good news is nothing so far has broken… well until later today.
One of the biggest changes is that we’re moving to a new, faster server.* In keeping with best tech practices, we’re going to do the final move, including redirecting that domain, on Friday afternoon.**
If all goes as planned OTB will become unavailable in the early afternoon, eastern time. We will then copy the most recent posts and comments over to the new instance and then, redirect our domain. If nothing scary happens, that should take about an hour and we should be up and running no later than 4pm at our new domain.***
To assist with the transition, I’ll be turning comments off at 1pm ET. That will make it easier to do the backup and minimize downtime. So set your watches and vent your spleens while you can!
*—Lesson learned: Always go back and check on old service contracts to see if you are paying for more than you are actually getting. Unfortunately, many companies will not tell you about deals that are better than the grandfathered plan you are on.
**—Ask me about the time in the late 1990’s I singlehandedly crashed the Kodak.com homepage and took it offline at 5:45pm on a summer Friday after most of the tech folks had gone home… wait, don’t bother, that’s pretty much the entire story.
***—For the time being there will be no visual or performance change between the old and new server. James and I decided to handle the migration before diving into some ongoing issues like the moderation problem some of you keep having.
Not to worry. There are enough cat videos on FB to keep me entertained for the duration.
(This guy might help you out in a jam.)
Right after lunch is an excellent time. There are support staff around (with your vendors) to help if it’s a disaster, no one misses lunch, and the new version will get enough traffic that you’re mostly confident it works.
Ask me about the time I followed my employer’s best practices, and did a launch during the 2am lull. I had to page in so many people, and we were all up until 5am sorting out why nothing worked. And then the entire thing blew up the next day in the early afternoon.
(The deployment problem was bad config, but when we fixed the bad config, it still failed and it turned out the deployment system itself was deployed with bad config at 1:50am. The outage the next day was caused by a changed SQL query that was missing an index — the DBAs only put the index on our staging DB. The COE (Cause/Correction of Error) report began “I’m sure everyone involved is normally very competent and meant well…”)
(My COEs were the best COEs.)
How will I ever demonstrate my outrage!!!!!!!!!!!!
C U on the other side…
It’s after one and I can still comment.
test
@Matt Bernius:..test
I C U
Well I thought I could easily disable comments… The migration has started so some comments are going to be lost… oh well.
Still accepting comments.
Just what do you think you are doing Dave?
@Mister Bluster:
Dave… Dave…
Hello, I am a HAL9000…