I'm looking to read a few (publicly available) postmortems of incidents and accidents. The domain/organization/... doesn't matter too much so long as they are insightful and there's something to be learned from reading them. Bonus points for including a quick note on why you considered them particularly interesting!
Chemical Safety Board has numerous casualties and fatalities:<p>www.csb.gov<p>Right there on the front page it says they will be investigating the recent H2S release that occurred at Pemex (the old Shell refinery) last week.<p>Other than that the most recent final report has been on the front page since July 2024, but that was an incident at LyondellBasell that occurred back in 2021, so these investigations are thorough but it takes time.<p>However within a day of the H2S fatalites at Pemex one of the county officials on live TV mentioned immediate hearsay that a "flange" of some type was a suspected culprit and contractors might be the most heavily involved more so than actual Pemex employees. Plenty of plant personnel were certainly exposed though, and it could be smelled across the neighborhoods way outside the gate. So it was not a minor leak, seems like some kind of reservoir was opened, and concentrated toxins emerged by the kilo and took the greatest toll on those in the immediate area.<p>If so that would be almost exactly like the LyondellBasell fatalities even though it was a different chemical.<p>IOW we may have to wait a few years to see if accurate conclusions can be made at Pemex, but it may very well turn out that contractors were performing <i>the exact same tasks</i>[0] at Pemex that occurred at LyondellBasell three years earlier. Probably using safety procedures well-established before 2021, but maybe not aware of this most recent CSB video. Which should really be required training material for anyone expected to be involved with these type tasks.<p>[0]Whether the potentially deadly tasks were being conducted carefully or carelessly would pretty much be a matter of judgment I think now.
<a href="https://www.pagerduty.com/blog/the-discovery-of-apache-zookeepers-poison-packet" rel="nofollow">https://www.pagerduty.com/blog/the-discovery-of-apache-zooke...</a> - an old postmortem that I still think was a masterclass in debugging, uncovering two bugs in ZooKeeper and two in the Linux kernel.<p><a href="https://about.gitlab.com/blog/2017/02/10/postmortem-of-database-outage-of-january-31" rel="nofollow">https://about.gitlab.com/blog/2017/02/10/postmortem-of-datab...</a> - thousands of GitLab projects permanently lost. A stark reminder of the importance of database backup testing procedures, because even multiple layers of backups can fail in a black swan event.
<a href="https://www.youtube.com/channel/UCXIkr0SRTnZO4_QpZozvCCA" rel="nofollow">https://www.youtube.com/channel/UCXIkr0SRTnZO4_QpZozvCCA</a><p>I am a massive follower of the USCSB. The videos have got the mix of retro early CGI that helps tell the story and can be a great aid in addition to reading the reports.
For incidents with Wikimedia's sites and infrastructure: <a href="https://wikitech.wikimedia.org/wiki/Incident_status" rel="nofollow">https://wikitech.wikimedia.org/wiki/Incident_status</a>
There's a repo with categories that gets updated every now and then<p><a href="https://github.com/danluu/post-mortems">https://github.com/danluu/post-mortems</a>