My CachyOS has worked flawlessly, but lately, I’m experiencing this problem:
❯ LANG=en-US free -h
total used free shared buff/cache available
Mem: 28Gi 13Gi 860Mi 549Mi 15Gi 14Gi
Swap: 54Gi 2.1Gi 52Gi
I have a lot of available memory (32 Gb of physical RAM)
I have zram0 swap enabled.
Some apps like ungoogled chromium and k3s are dying because of low memory.
k3s is a kubernetes cluster run locally.
In dmseg I can see some messages like this:
[15476.990366] node invoked oom-killer: gfp_mask=0x140dca(GFP_HIGHUSER_MOVABLE|__GFP_COMP|__GFP_ZERO), order=0, oom_score_adj=1000
[15476.990375] CPU: 13 UID: 0 PID: 126762 Comm: node Not tainted 6.14.0-4-cachyos #1 016ff1edc592562531da7158aff85145022cf30c
[15476.990378] Hardware name: Trigkey S5/S5, BIOS 5800H506 03/29/2023
[15476.990380] Call Trace:
[15476.990382] <TASK>
[15476.990386] dump_stack_lvl+0x71/0x90
[15476.990391] dump_header+0x50/0x1f0
[15476.990394] oom_kill_process+0x1db/0x2c0
[15476.990397] out_of_memory+0x4ab/0x5f0
[15476.990399] __alloc_pages_slowpath+0x729/0x1050
[15476.990405] ? __alloc_frozen_pages_noprof.cold+0x96/0xf9
[15476.990409] ? vma_alloc_folio_noprof+0x12a/0x210
[15476.990412] ? do_pte_missing+0xde/0x12e0
[15476.990414] ? srso_alias_return_thunk+0x5/0xfbef5
[15476.990415] ? __mod_memcg_lruvec_state.llvm.1293502753125321087+0x8e/0x190
[15476.990418] ? srso_alias_return_thunk+0x5/0xfbef5
[15476.990420] ? handle_mm_fault+0x536/0x850
[15476.990425] ? do_user_addr_fault+0x1c3/0x410
[15476.990428] ? exc_page_fault+0x7d/0x95
[15476.990431] ? asm_exc_page_fault+0x26/0x30
[15476.990436] </TASK>
[15476.990437] Mem-Info:
[15476.990441] active_anon:2255074 inactive_anon:737983 isolated_anon:0
active_file:506056 inactive_file:2961309 isolated_file:0
unevictable:117 dirty:1927 writeback:0
slab_reclaimable:328230 slab_unreclaimable:108811
mapped:664268 shmem:127372 pagetables:44954
sec_pagetables:1205 bounce:0
kernel_misc_reclaimable:0
free:45370 free_pcp:8402 free_cma:0
swap config:
❯ swapon -s
Nome de arquivo Tipo Tamanho Usado Prioridade
/dev/zram0 partition 29680636 2187000 100
/dev/nvme0n1p1 partition 27339772 34316 -2
Kernel parameters:
❯ cat /boot/refind_linux.conf
"Boot with standard options" "quiet zswap.enabled=0 nowatchdog splash splash rw rootflags=subvol=/@ root=UUID=29b63aea-88f6-4bc6-b3cb-698f1a36d4d7 resume=UUID=df119baa-39a0-4654-a9ef-51f1e03f05f3"
"Boot to single-user mode" "quiet zswap.enabled=0 nowatchdog splash splash rw rootflags=subvol=/@ root=UUID=29b63aea-88f6-4bc6-b3cb-698f1a36d4d7" single
"Boot with minimal options" "ro root=/dev/nvme0n1p3"
I have taken some measures like reducing the size of tmpfs to 2 Gb, but no luck in resolving my problem.