0

My cardano-node is failing, and I'm not sure how to debug or fix it:

NodeToClientV_10    HardForkNodeToClientEnabled HardForkSpecificNodeToClientVersion2 (EraNodeToClientEnabled ByronNodeToClientVersion1 :* EraNodeToClientEnabled ShelleyNodeToClientVersion4 :* EraNodeToClientEnabled ShelleyNodeToClientVersion4 :* EraNodeToClientEnabled ShelleyNodeToClientVersion4 :* EraNodeToClientEnabled ShelleyNodeToClientVersion4 :* Nil)
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:44:25.51 UTC] Opened imm db with immutable tip at 1ae2812fe7d76761e9d6f3733fcb8ee9c468040152ae8ad7f58d0527703e443f at slot 51967021 and chunk 2405
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:44:33.05 UTC] Opened vol db
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:25.71 UTC] Replaying ledger from snapshot DiskSnapshot {dsNumber = 51964788, dsSuffix = Nothing} at 8f134f487815d32650107fb04609ad1a7fbab47daaa55d27e6a4b4c3588c705f at slot 51964788
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:26.22 UTC] Replayed block: slot SlotNo 51964807 of At (SlotNo 51967021)
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:37.13 UTC] block replay progress (%) = 100.0
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:38.79 UTC] before next, messages elided = 51964826
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:38.79 UTC] Replayed block: slot SlotNo 51967021 of At (SlotNo 51967021)
[vvv-:cardano.node.ChainDB:Info:5] [2022-02-01 02:45:38.79 UTC] Opened lgr db
Killed
1
  • Yes, it looks like it was out of memory: [102938.126815] Out of memory: Killed process 28026 (cardano-node) total-vm:1074350476kB, anon-rss:7832444kB, file-rss:2160kB, shmem-rss:0kB, UID:1000 pgtables:15576kB oom_score_adj:0 [102938.346596] oom_reaper: reaped process 28026 (cardano-node), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB Feb 1, 2022 at 3:34

1 Answer 1

2

Killed probably means it was killed by the OOM (Out Of Memory) killer. If it was OOM killed, there will be a message in the dmesg output.

Running the node requires at least 8GB and you should have 4G of swap as well.

1
  • journalctl -k around that time stamp is what you'd run to see kernel logs of what was killed. Feb 1, 2022 at 13:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.