nimbus-eth1
nimbus-eth1 copied to clipboard
Nimbus does not stop when container is stopped
docker-compose sends a SIGTERM by default, and Nimbus does not appear to react to it. This is Nimbus in a container, with an entrypoint script that then calls exec "$@"
to transfer control to Nimbus. Note the code 137 when docker-compose finally force-kills Nimbus
eth-docker-devel-execution-1 | INF 2022-12-03 19:25:38.163+00:00 Forkchoice requested sync to new head tid=1 file=engine_api.nim:238 number=16106194 hash=B0DABE02969BEC730E765A3CCFA04BF2AF526560E2D7E181C789A3348D7DCAE0
eth-docker-devel-execution-1 | WRN 2022-12-03 19:25:38.320+00:00 Ignoring payload with missing parent tid=1 file=engine_api.nim:97 number=16106196 hash=5F65A1A7FED7142B068832BE3D7DADAA661201B7287AFFFFB279E2AAEBC053A8 parent=119EC9AEEE1F4F4316BE996AB4488EE73AF4A1EF51CB8FF4D62D8D1A8AA1423F
eth-docker-devel-execution-1 | INF 2022-12-03 19:25:38.355+00:00 Forkchoice requested sync to new head tid=1 file=engine_api.nim:238 number=16106195 hash=119EC9AEEE1F4F4316BE996AB4488EE73AF4A1EF51CB8FF4D62D8D1A8AA1423F
eth-docker-devel-execution-1 | INF 2022-12-03 19:25:38.662+00:00 Forkchoice requested sync to new head tid=1 file=engine_api.nim:238 number=16106196 hash=5F65A1A7FED7142B068832BE3D7DADAA661201B7287AFFFFB279E2AAEBC053A8
eth-docker-devel-execution-1 | INF 2022-12-03 19:26:01.723+00:00 Snap sync statistics topics="snap-tick" tid=1 file=ticker.nim:153 up=0 buddies=1 pivot=n/a nAcc=0(0) accCov=0%(0%)/0%~0/0 nSto=0(0) nStoQue=n/a mem=11.30m
eth-docker-devel-execution-1 | INF 2022-12-03 19:26:02.033+00:00 Snap pivot initialised topics="snap-buddy" tid=1 file=worker.nim:136 peer=Node[144.76.29.49:30303] pivot=#15050010 multiOk=false runState=Running
eth-docker-devel-execution-1 | INF 2022-12-03 19:26:02.724+00:00 Snap sync statistics topics="snap-tick" tid=1 file=ticker.nim:153 up=1 buddies=1 pivot=#15050010/1 nAcc=0(0) accCov=0%(0%)/0%~0/1 nSto=0(0) nStoQue=0 mem=12.46m
eth-docker-devel-execution-1 exited with code 137
If I add stop_signal: SIGINT
then Nimbus will stop as expected. It seems like it's just not responding to SIGTERM
but does to SIGINT