We've had two recent issues:
- downtime whenever we release new code, leading to 502 errors
- an incident where we served very slow requests and sometimes leading to 502 errors
We've fixed these issues. Now we have zero downtime deployments again. We migrated to a new server about a year ago, which came with the loss of zero downtime deployments. We've worked around it for this long, but it became a big enough problem for Bookhead users and ourselves that we've solved this problem.
We also released a fix that addressed the second issue with slow requests, which was unrelated to the deployments. This was an engineering optimization we needed to make.
See the uptime status
We have a website that monitors our systems and alerts us whenever there is an outage. It also gives a general status of our services. If you are ever experiencing an issue where you can't access Bookhead, then the best place to check for updates is our status page:
If you see all the red in the recent days, those were the problem days! It's kind of like your bookstore's internet going down like 1-2 times a day while your store is open. It looks like the service is down those entire days, but a day will become red even if it's down for 5 minutes. Our reliability stats are better than some of the software companies that are large corporations. We feel like we are doing a good job at it! But can always do better.