RPI3 swap experiments (grace under pressure)

[Jedi Tek'Unum]

I firmly disagree with the entire concept of out of memory killers. They are simply evil and in my opinion a complete cop-out. I first encountered this kind of kludge back in the ‘80s with AIX. It was bad then and it still is today. Frankly I find it ridiculous that they still exist.

For 35+ years I’ve worked with many unix variants on everything from supercomputers on down. A majority on SunOS/Solaris. BIG Solaris systems running crushing workloads.

I have one question… when was the last time anyone saw Solaris kill a process because the system was under memory stress? In my experience, NEVER! And I wouldn’t say that the system became unreasonably unresponsive either.

As far as I’m concerned, any system deployed in a “mission critical role” (and I’m not referring to life-critical) has no business ever killing a process for load reasons. Period.

I’d go further and say that any unix system today has had plenty of time to evolve into a “utility-grade” service and is therefore expected to support exactly that kind of mission critical role. To claim that it isn’t possible, or practical, just doesn’t hold water - Solaris has been doing it for a LONG TIME.

I’m fully aware of the disaster related to the current owner of Solaris. I’m not advocating the owner or, unfortunately, the technology for the same reason. But that doesn’t change the technical basis of saying this is a level of engineering that is expected.