Post #1546563
2026-04-14 17:50 UTC
@cliffle @regehr I have a conjecture that almost all of Linux's problems with swap thrashing and OOM killing of random processes are down to two bad defaults: memory overcommit and allowing 30% of phys RAM to fill up with dirty pages before it even starts *background* writeback.
Replies (2)
-
@zwol@masto.hackers.town 2026-04-14 17:56
@cliffle @regehr so there's not nearly enough back pressure on processes generating dirty pages, and the dirty pages choke out everything else
-
@cliffle@hachyderm.io 2026-04-14 22:02
@zwol @regehr while I sort of agree with parts of that, I wanted to point out that disabling memory overcommit *also* invokes a regulatory system without enough context: the allocator. Whether two programs can succeed now depends on their execution interleaving of their map requests, and the decision to fail must be made even earlier (at the time of mapping). So while I agree it's often an improvement, it doesn't totally fix the issue I was describing.