Graviton3 support – intermittent crash/coredump

What's going wrong?

• pm2 daemon crashes (coredump logged in /var/log/messages)

How could we reproduce this issue?

• it doesn't seem to reproduce consistently. maybe 10% of our deploys (pm2 startOrReload) result in a crash/coredump.
• We just moved from graviton2 to graviton3, so I suspect that the issue is related to an incompatibility with graviton3.

Supporting information

--- PM2 report ----------------------------------------------------------------
Date                 : Mon Apr 08 2024 18:32:37 GMT+0000 (Coordinated Universal Time)
===============================================================================
--- Daemon -------------------------------------------------
pm2d version         : 5.3.0
node version         : 18.17.0
node path            : /home/ec2-user/.nvm/versions/node/v18.17.0/bin/pm2
argv                 : /home/ec2-user/.nvm/versions/node/v18.17.0/bin/node,/home/ec2-user/.nvm/versions/node/v18.17.0/lib/node_modules/pm2/lib/Daemon.js
argv0                : node
user                 : ec2-user
uid                  : 1000
gid                  : 1000
uptime               : 57min
===============================================================================
--- CLI ----------------------------------------------------
local pm2            : 5.3.0
node version         : 18.17.0
node path            : /home/ec2-user/.nvm/versions/node/v18.17.0/bin/pm2
argv                 : /home/ec2-user/.nvm/versions/node/v18.17.0/bin/node,/home/ec2-user/.nvm/versions/node/v18.17.0/bin/pm2,report
argv0                : node
user                 : ec2-user
uid                  : 1000
gid                  : 1000
===============================================================================
--- System info --------------------------------------------
arch                 : arm64
platform             : linux
type                 : Linux
cpus                 : unknown
cpus nb              : 4
freemem              : 14946652160
totalmem             : 16449142784
home                 : /home/ec2-user
===============================================================================
--- PM2 list -----------------------------------------------
┌────┬───────────┬─────────────┬─────────┬─────────┬──────────┬────────┬──────┬───────────┬──────────┬──────────┬──────────┬──────────┐
│ id │ name      │ namespace   │ version │ mode    │ pid      │ uptime │ ↺    │ status    │ cpu      │ mem      │ user     │ watching │
└────┴───────────┴─────────────┴─────────┴─────────┴──────────┴────────┴──────┴───────────┴──────────┴──────────┴──────────┴──────────┘
===============================================================================
--- Daemon logs --------------------------------------------
/home/ec2-user/.pm2/pm2.log last 20 lines:
PM2        | 2024-04-08T17:26:16: PM2 log: pid=149369 msg=failed to kill - retrying in 100ms
PM2        | 2024-04-08T17:26:16: PM2 log: pid=137480 msg=failed to kill - retrying in 100ms
PM2        | 2024-04-08T17:26:16: PM2 log: Process with pid 149369 still alive after 30000ms, sending it SIGKILL now...
PM2        | 2024-04-08T17:26:17: PM2 log: Process with pid 137480 still alive after 30000ms, sending it SIGKILL now...
PM2        | 2024-04-08T17:34:44: PM2 log: ===============================================================================
PM2        | 2024-04-08T17:34:45: PM2 log: --- New PM2 Daemon started ----------------------------------------------------
PM2        | 2024-04-08T17:34:45: PM2 log: Time                 : Mon Apr 08 2024 17:34:45 GMT+0000 (Coordinated Universal Time)
PM2        | 2024-04-08T17:34:45: PM2 log: PM2 version          : 5.3.0
PM2        | 2024-04-08T17:34:45: PM2 log: Node.js version      : 18.17.0
PM2        | 2024-04-08T17:34:45: PM2 log: Current arch         : arm64
PM2        | 2024-04-08T17:34:45: PM2 log: PM2 home             : /home/ec2-user/.pm2
PM2        | 2024-04-08T17:34:45: PM2 log: PM2 PID file         : /home/ec2-user/.pm2/pm2.pid
PM2        | 2024-04-08T17:34:45: PM2 log: RPC socket file      : /home/ec2-user/.pm2/rpc.sock
PM2        | 2024-04-08T17:34:45: PM2 log: BUS socket file      : /home/ec2-user/.pm2/pub.sock
PM2        | 2024-04-08T17:34:45: PM2 log: Application log path : /home/ec2-user/.pm2/logs
PM2        | 2024-04-08T17:34:45: PM2 log: Worker Interval      : 30000
PM2        | 2024-04-08T17:34:45: PM2 log: Process dump file    : /home/ec2-user/.pm2/dump.pm2
PM2        | 2024-04-08T17:34:45: PM2 log: Concurrent actions   : 2
PM2        | 2024-04-08T17:34:45: PM2 log: SIGTERM timeout      : 1600
PM2        | 2024-04-08T17:34:45: PM2 log: ===============================================================================

Start by investigating the intermittent crash during pm2 startOrReload on Graviton3, using the provided PM2 report, daemon logs, and /var/log/messages coredump records. Compare the reported PM2 and Node.js versions with the arm64 environment; done means the failure is reproducible or its cause is isolated and the daemon no longer crashes during deployment.