store: judge memory pressure by swap RATE, not level #130
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: neuron-technologies/el#130
Reference in New Issue
Block a user
Delete Branch "fix/elc-rebuildable-compiler-builtins"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
The guard added minutes earlier checked swap as a level (
avail < total/8). Wrong signal — the same host proved it twice within minutes:Both are ~97% "used". macOS grows swap on demand and trims lazily, so the level is a high-water mark that says almost nothing about now. A level check calls the second state an emergency and starves the pool for no reason — and a guard that fires on healthy machines gets disabled, and then guards nothing.
What separates them is whether pages are moving. Now samples the swapout counter across calls and judges the delta:
Thresholds measured, not guessed: idle 0/s; recovery burst 24,845/s while the compressor drained (transient, and correctly not a growth decision since growth is only evaluated on eviction passes); real thrash ~2000/s. 200/s sits clearly above noise and far below either.
Compressor-footprint subtraction stays — that RAM is spoken for regardless of paging rate.
The guard I added minutes ago checked swap availability as a level (avail < total/8 -> report zero available). That is the wrong signal, and the same host proved it twice within minutes: 47.65 / 48.00 GiB swap used, 2047 swapouts/s -> genuinely thrashing 26.67 / 28.00 GiB swap used, 0 swapouts/s -> healthy, 15.6 GiB free Both are ~97% "used". macOS grows swap files on demand and trims them lazily, so the level says almost nothing about now — it is a high-water mark. The level check calls the second state an emergency and starves the pool for no reason, which is its own failure mode: a guard that fires on healthy machines gets disabled, and then guards nothing. What separates the two is whether pages are moving. So sample the swapout counter across calls and judge the delta: - > 200 pages/s (~3 MiB/s) sustained outward paging => report zero available; callers refuse to grow and pc_relieve_pressure hands frames back. - The first call primes the baseline and reports no pressure. One sample cannot have a rate, and inferring one from a single reading is exactly the mistake this commit removes. Measured thresholds, not guessed: idle sat at 0/s, recovery burst hit 24,845/s while the compressor drained (transient, correctly not a growth decision since growth is only evaluated on eviction passes), and real thrash held ~2000/s. 200/s sits clearly above noise and far below either. The compressor-footprint subtraction stays: that RAM is genuinely spoken for regardless of paging rate.