0

We have a server runnnig Linux 5.15 and we've had multiple verified situations where a process is killed by OOM and that leaves the whole system inaccessible on the network, both for inbound and outbound traffic. This is a recent syslog trail for the event:

Mar  8 05:16:01 ip-10-110-10-133 kernel: [203986.004138] amazon-cloudwat invoked oom-killer: gfp_mask=0x1100cca(GFP_HIGHUSER_MOVABLE), order=0, oom_score_adj=0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004146] CPU: 3 PID: 1627 Comm: amazon-cloudwat Not tainted 5.15.0-1031-aws #35~20.04.1-Ubuntu
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004149] Hardware name: Amazon EC2 r6i.2xlarge/, BIOS 1.0 10/16/2017
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004150] Call Trace:
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004152]  <TASK>
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004155]  dump_stack_lvl+0x4a/0x63
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004162]  dump_stack+0x10/0x16
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004164]  dump_header+0x53/0x225
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004168]  oom_kill_process.cold+0xb/0x10
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004171]  out_of_memory+0x1dc/0x530
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004174]  __alloc_pages_slowpath.constprop.0+0xd32/0xe30
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004179]  ? __alloc_pages_slowpath.constprop.0+0xdb6/0xe30
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004182]  __alloc_pages+0x2cc/0x310
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004185]  alloc_pages+0x90/0x120
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004187]  __page_cache_alloc+0x87/0xc0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004191]  pagecache_get_page+0x150/0x530
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004193]  ? page_cache_ra_unbounded+0x16a/0x220
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004196]  filemap_fault+0x527/0xb60
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004198]  ? filemap_map_pages+0x138/0x640
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004201]  __do_fault+0x3d/0x120
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004205]  do_fault+0x1f9/0x420
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004207]  __handle_mm_fault+0x62c/0x840
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004210]  handle_mm_fault+0xd8/0x2c0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004211]  do_user_addr_fault+0x1c2/0x660
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004215]  exc_page_fault+0x77/0x170
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004218]  asm_exc_page_fault+0x27/0x30
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004221] RIP: 0033:0x44c1a0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004235] Code: Unable to access opcode bytes at RIP 0x44c176.
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004235] RSP: 002b:000000c001705de8 EFLAGS: 00010246
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004238] RAX: 000000c001705f78 RBX: 000000c001705e8c RCX: 0000000000000000
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004240] RDX: 0000000000000000 RSI: 0000000000000003 RDI: 0000000000000000
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004241] RBP: 000000c001705fb8 R08: 0000000000000001 R09: 000000c00063bb30
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004242] R10: 000000c001705f00 R11: 000000c000b715c0 R12: 000000c000d73080
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004243] R13: ffffffffffffffff R14: 000000c000bec820 R15: 0000000000000000
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004246]  </TASK>
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004247] Mem-Info:
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248] active_anon:202 inactive_anon:15952689 isolated_anon:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  active_file:146 inactive_file:0 isolated_file:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  unevictable:6279 dirty:3 writeback:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  slab_reclaimable:9465 slab_unreclaimable:13692
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  mapped:68073 shmem:256 pagetables:33680 bounce:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  kernel_misc_reclaimable:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004248]  free:91381 free_pcp:1688 free_cma:0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004253] Node 0 active_anon:808kB inactive_anon:63810756kB active_file:584kB inactive_file:0kB unevictable:25116kB isolated(anon):0kB isolated(file):0kB mapped:272292kB dirty:12kB writeback:0kB shmem:1024kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 0kB writeback_tmp:0kB kernel_stack:5536kB pagetables:134720kB all_unreclaimable? no
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004257] Node 0 DMA free:11264kB min:16kB low:28kB high:40kB reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB present:15992kB managed:15360kB mlocked:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004262] lowmem_reserve[]: 0 2991 63273 63273 63273
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004266] Node 0 DMA32 free:244188kB min:3188kB low:6244kB high:9300kB reserved_highatomic:0KB active_anon:0kB inactive_anon:2804896kB active_file:0kB inactive_file:544kB unevictable:0kB writepending:0kB present:3129252kB managed:3063716kB mlocked:0kB bounce:0kB free_pcp:684kB local_pcp:112kB free_cma:0kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004270] lowmem_reserve[]: 0 0 60281 60281 60281
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004273] Node 0 Normal free:110072kB min:115576kB low:177292kB high:239008kB reserved_highatomic:2048KB active_anon:808kB inactive_anon:61005860kB active_file:1364kB inactive_file:388kB unevictable:25116kB writepending:12kB present:62898176kB managed:61728412kB mlocked:18340kB bounce:0kB free_pcp:6084kB local_pcp:928kB free_cma:0kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004278] lowmem_reserve[]: 0 0 0 0 0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004281] Node 0 DMA: 0*4kB 0*8kB 0*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 1*1024kB (U) 1*2048kB (M) 2*4096kB (M) = 11264kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004292] Node 0 DMA32: 188*4kB (UME) 127*8kB (UME) 135*16kB (UME) 69*32kB (UME) 33*64kB (UME) 12*128kB (UME) 8*256kB (UME) 2*512kB (UE) 2*1024kB (UM) 2*2048kB (ME) 55*4096kB (M) = 244280kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004306] Node 0 Normal: 9199*4kB (UME) 5601*8kB (UME) 1449*16kB (UMEH) 193*32kB (UMEH) 7*64kB (MH) 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 111412kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004318] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004320] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004321] 4359 total pagecache pages
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004322] 0 pages in swap cache
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004323] Swap cache stats: add 0, delete 0, find 0/0
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004324] Free swap  = 0kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004325] Total swap = 0kB
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004326] 16510855 pages RAM
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004326] 0 pages HighMem/MovableOnly
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004327] 308983 pages reserved
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004328] 0 pages hwpoisoned
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004328] Tasks state (memory values in pages):
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004329] [  pid  ]   uid  tgid total_vm      rss pgtables_bytes swapents oom_score_adj name
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004337] [    211]     0   211   106563     1109   811008        0          -250 systemd-journal
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004340] [    248]     0   248     2247      979    61440        0         -1000 systemd-udevd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004343] [    340]     0   340    53652     4488    94208        0         -1000 multipathd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004346] [    360]     0   360      701       29    45056        0             0 falcond
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004348] [    361]     0   361   696911   147512  1540096        0             0 falcon-sensor-b
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004350] [    370]     0   370     2841      375    45056        0         -1000 auditd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004352] [    441]   100   441     6680     1009    77824        0             0 systemd-network
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004355] [    446]   101   446     6001     1658    86016        0             0 systemd-resolve
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004357] [    534]     0   534    60348      354   102400        0             0 accounts-daemon
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004359] [    535]     0   535      637      165    40960        0             0 acpid
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004361] [    536]     0   536   200300    13159   385024        0             0 amazon-cloudwat
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004363] [    540]   103   540     1920      952    53248        0          -900 dbus-daemon
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004365] [    568]     0   568    20476      612    61440        0             0 irqbalance
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004367] [    570]   113   570     3256      342    53248        0             0 chronyd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004369] [    577]     0   577     7494     2846    98304        0             0 networkd-dispat
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004370] [    580]   113   580     1210      439    53248        0             0 chronyd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004372] [    603]     0   603     2168      585    57344        0             0 cron
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004375] [    609]     0   609    59107      779    94208        0             0 polkitd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004377] [    617]   104   617    56125      866    86016        0             0 rsyslogd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004379] [    621]     0   621    12231     6333   126976        0             0 salt-minion
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004381] [    628]     0   628     4307     1002    69632        0             0 systemd-logind
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004382] [    632]     0   632    98669      829   131072        0             0 udisksd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004384] [    636]     0   636      951      499    49152        0             0 atd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004386] [    677]     0   677     1840      447    53248        0             0 agetty
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004388] [    680]     0   680     4561      533    61440        0             0 wrapper
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004389] [    700]     0   700     3047      932    61440        0         -1000 sshd
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004391] [    735]     0   735    60152     1054   102400        0             0 ModemManager
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004393] [    736]     0   736     1459      385    49152        0             0 agetty
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004395] [    737]     0   737    27031     2719   110592        0             0 unattended-upgr
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004397] [    857]     0   857  1461116   107789  1564672        0             0 java
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004399] [    928]     0   928   248205    12742   299008        0             0 salt-minion
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004401] [   1041]     0  1041    31500     6490   143360        0             0 salt-minion
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004403] [   1082]     0  1082     9519      580    69632        0             0 master
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004405] [   1084]   112  1084     9670      566    65536        0             0 qmgr
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004407] [   3599]   112  3599    10536      724    69632        0             0 tlsmgr
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004409] [  97274]     0 97274   307044      307   163840        0             0 newrelic-infra-
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004411] [  97282]     0 97282   440719     4508   294912        0             0 newrelic-infra
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004413] [ 251487]   112 251487     9585      121    65536        0             0 pickup
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004416] [ 257287]     0 257287     2553      624    57344        0             0 cron
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004418] [ 257288]     0 257288     2553      624    57344        0             0 cron
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004419] [ 257289]     0 257289     2553      623    57344        0             0 cron
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004421] [ 257290]     0 257290     2553      623    57344        0             0 cron
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004423] [ 257292]  3001 257292     2189      118    49152        0             0 bash
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004425] [ 257293]  3001 257293     2189      115    57344        0             0 bash
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004427] [ 257294]  3001 257294     2189      119    49152        0             0 bash
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004429] [ 257296]  3001 257296     2189      118    57344        0             0 bash
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004431] [ 257307]  3001 257307      656       29    40960        0             0 run-one
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004433] [ 257308]  3001 257308      656       29    40960        0             0 run-one
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004434] [ 257309]  3001 257309      656       29    40960        0             0 run-one
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004436] [ 257310]  3001 257310      656       29    40960        0             0 run-one
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004438] [ 257344]  3001 257344     1859       24    53248        0             0 flock
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004440] [ 257346]  3001 257346  7911996  7731555 62894080        0             0 python
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004442] [ 257347]  3001 257347     1859       24    61440        0             0 flock
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004443] [ 257348]  3001 257348     1859       24    49152        0             0 flock
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004445] [ 257349]  3001 257349   338380   158223  2191360        0             0 python
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004447] [ 257351]  3001 257351   365119   184973  2453504        0             0 python
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004449] [ 257355]  3001 257355     1859       25    53248        0             0 flock
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004451] [ 257356]  3001 257356  7883801  7703167 62664704        0             0 python
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004452] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/system.slice/cron.service,task=python,pid=257346,uid=3001
Mar  8 05:16:02 ip-10-110-10-133 kernel: [203986.004468] Out of memory: Killed process 257346 (python) total-vm:31647984kB, anon-rss:30926220kB, file-rss:0kB, shmem-rss:0kB, UID:3001 pgtables:61420kB oom_score_adj:0
Mar  8 05:17:01 ip-10-110-10-133 CRON[258623]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Mar  8 05:19:31 ip-10-110-10-133 postfix/smtp[258801]: D826A400AE: to=<[email protected]>, relay=email-smtp.us-east-1.amazonaws.com[52.20.44.183]:587, delay=0.39, delays=0.01/0.02/0.13/0.23, dsn=2.0.0, status=sent (250 Ok 01000186bfa906f8-4a6fed17-2ceb-423d-8476-e161c40962b0-000000)
Mar  8 05:19:31 ip-10-110-10-133 postfix/smtp[258801]: D826A400AE: to=<[email protected]>, relay=email-smtp.us-east-1.amazonaws.com[52.20.44.183]:587, delay=0.39, delays=0.01/0.02/0.13/0.23, dsn=2.0.0, status=sent (250 Ok 01000186bfa906f8-4a6fed17-2ceb-423d-8476-e161c40962b0-000000)
Mar  8 05:19:31 ip-10-110-10-133 postfix/smtp[258801]: D826A400AE: to=<[email protected]>, relay=email-smtp.us-east-1.amazonaws.com[52.20.44.183]:587, delay=0.39, delays=0.01/0.02/0.13/0.23, dsn=2.0.0, status=sent (250 Ok 01000186bfa906f8-4a6fed17-2ceb-423d-8476-e161c40962b0-000000)
Mar  8 05:23:03 ip-10-110-10-133 newrelic-infra-service[97282]: time="2023-03-08T05:23:03Z" level=warning msg="commands poll failed" component=CommandChannelService error="command request submission failed: Get \"https://infrastructure-command-api.newrelic.com/agent_commands/v1/commands\": context deadline exceeded (Client.Timeout exceeded while awaiting headers)"
Mar  8 05:27:24 ip-10-110-10-133 newrelic-infra-service[97282]: time="2023-03-08T05:27:24Z" level=warning msg="error occurred while updating the system fingerprint" component=Agent error="unable to fetch AWS metadata: Get \"http://169.254.169.254/latest/dynamic/instance-identity/document\": dial tcp 169.254.169.254:80: i/o timeout"
Mar  8 05:29:29 ip-10-110-10-133 newrelic-infra-service[97282]: time="2023-03-08T05:27:24Z" level=warning msg="commands poll failed" component=CommandChannelService error="command request submission failed: Get \"https://infrastructure-command-api.newrelic.com/agent_commands/v1/commands\": dial tcp 162.247.242.49:443: i/o timeout (Client.Timeout exceeded while awaiting headers)"
Mar  8 05:34:38 ip-10-110-10-133 systemd[1]: Starting Ubuntu Advantage Timer for running repeated jobs...
Mar  8 06:01:58 ip-10-110-10-133 systemd[1]: collector.service: Main process exited, code=exited, status=1/FAILURE
Mar  8 06:07:08 ip-10-110-10-133 systemd[1]: collector.service: Failed with result 'exit-code'.
Mar  8 06:09:58 ip-10-110-10-133 systemd-networkd[441]: ens5: Could not set DHCPv4 address: Connection timed out
Mar  8 06:12:04 ip-10-110-10-133 systemd-networkd[441]: ens5: Failed

Right after the OOM, a lot of services start failing, in a way that looks like it's not just DNS. Why would this be the case? Shouldn't the killed process be the only one affected in an OOM?

Also, what's puzzling is that somehow smtpd prints a success message in relaying an email after the incident -- I'm not sure if this is a red herring or not, but all other services report network errors after the OOM. A reboot of course fixes all issues.

1 Answer 1

0

I'm not convinced that OOM killing some python process is what caused IPv4 problems. Especially something as presumably as unrelated as some Python script started from cron.

An alternate explanation is those are symptoms of different faults. Why indeed would there be SMTP messages successfully relayed, a couple minutes after the OOM?

Your system could be performing horribly bad under such memory pressure. Sufficiently bad that IP connections fail and services stop working effectively would be extraordinary, but possible. Look at any performance metrics available. Load average numbers should be not to insane for a box with this number of CPUs. Memory pressure stall information would be very valuable, to prove if any time was spent thrashing on memory. Running processes suggest you are using AWS CloudWwatch or New Relic, use those application or host metrics as well.

On the other hand, it is bad for system stability and application correctness for some tasks to be forcibly killed. Even if usually the system recovers, the effects of killing the wrong application or daemon or agent could be bad. Linux virtual memory system tries very hard not to OOM kill, its one of the last resort actions.

If the workloads on this box were doing reasonable things, look at capacity planning, memory might not be sufficient.

2
  • 1
    We have experienced this on AWS EC2 when we run out of the memory and then DHCP lease is not refreshed correctly (hence the systemd-networkd[38912]: ens5: Failed error) and expires after a while. As a result, the server has no IP address and stops communicating.
    – Ondrej
    May 2, 2023 at 11:43
  • Should a DHCP client dying be a risk of networking going down, there are alternatives. In particular, IPv6 with SLAAC avoids DHCP entirely. However that doesn't fix the root cause of being unprepared for a 30 GB memory process. May 10, 2023 at 14:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .