The rest of the story explains how they isolated the problem and worked around it; it turns out that the culprit was TRIM, or rather TRIM's interaction with certain SSDs: "The system was issuing a TRIM to erase empty blocks, the command got misinterpreted by the drive and the controller erased blocks it was not supposed to. Therefore our files ended-up with 512 bytes of zeroes, files smaller than 512 bytes were completely zeroed. When we were lucky enough, the misbehaving TRIM hit the super-block of the filesystem and caused a corruption."
Since SSDs are becoming the norm outside the data center as well as within, some of the problems that their analysis exposed for one company probably would be good to test for elsewhere. One upshot: "As a result, we informed our server provider about the affected SSDs and they informed the manufacturer. Our new deployments were switched to different SSD drives and we don't recommend anyone to use any SSD that is anyhow mentioned in a bad way by the Linux kernel."
Epoch uses a numeric priority system to determine the boot order of objects, supports a wide range of customizability, and doesn't require much anything except libc and /bin/sh (though /bin/sh can be omitted, not recommended). Epoch also features advanced service status reporting features and has service supervision.
I'm just here to ask the Slashdot community what they'd like to see in the next release, 1.3.0 "Fluoxetine", before I wrap it up.
There are generally 2 things I can't consider:
* Parallel service startup, because that can be done manually, and implementing it would make Epoch too complex IMHO.
* Switching away from the numeric priority system to "true" dependencies. I implemented the priority system because I liked the freedom it gives the end user.
Despite these omissions, your feedback matters to me. I want to make something everyone will want to use.
These videos (both visible today) were made remotely, with Timothy Lord at one end in Austin, TX, and Jim Zemlin at the other end in Tokyo, Japan. Their sound quality suffers from the distance involved, but they are generally intelligible -- and, of course, you can always choose to read the transcript instead of watching the videos.