Accidentally deleted Gitlab's production database
Images
One night, a software developer working in the Netherlands accidentally deleted Gitlab's production database while trying to debug some issues. He did this by entering the delete all command in the wrong terminal. In trying to fix his delete mistake, more problems were uncovered; the database backups were not working as expected. Luckily, this developer took a snapshot of the database six hours before. They would use the snapshot to recover as much data as possible. It would take 18+ hours to recover fully, and Gitlab would lose six hours of merge requests and other metadata.
The company handled this incident well, sharing out the recovery process. They would not fire the developer. The developer had around 6.5 years of programming experience at that point, but he only joined Gitlab a little more than a year before the mistake. He would stay at the company for nearly five more years before leaving of his own accord. GitLab has IPOed since then and has over 30 million users today.
Social Media
Quotes
GitLab handled this very well,...Nobody got fired or yelled at, everybody realised this was a problem with the organisation as a whole.
Yorick Peterse
Hi, guy here who accidentally nuked GitLab.com's database earlier this year. Fortunately we did have a backup, though it was 6 hours old at that point.
Yorick Peterse
References
-
GitHub vs GitLab: Which is the Best in 2023? Differences and Similarities
Considering the popularity between GitHub and GitLab, GitHub has around 56 million users, whereas GitLab has over 31 million users. But GitLab has its own niche.
Radix Web
-
I'm leaving GitLab to work on Inko full-time
Back in October 2015 I joined GitLab. I think I was employee #28 at the time, with the total number of employees being somewhere between 30 and 40 if I'm not mistaken. Fast-forward to today, and GitLab has grown to almost 1600 employees.
yorickpeterse
-
The tech world is rallying around a young developer who made a huge, embarrassing mistake
GitLab chronicled its recovery efforts live on YouTube and in a Google doc, and treated it as a company problem instead of an individual one.
Yahoo
-
Postmortem of database outage of January 31
Specifically, we lost modifications to database data such as projects, comments, user accounts, issues and snippets, that took place between 17:20 and 00:00 UTC on January 31.
Gitlab
-
GitLab.com melts down after wrong directory deleted, backups fail
Behind the scenes, a tired sysadmin, working late at night in the Netherlands, had accidentally deleted a directory on the wrong server during a frustrating database replication process: he wiped a folder containing 300GB of live production data that was due to be replicated.
TheRegister
-
GitLab suffers major backup failure after data deletion incident
One positive thing that can be said is the company is at least providing a high level of detail about the incident — however cold comfort that is if you’re facing the loss of a bunch of data.
TechCrunch