High Resource usage on v24.11.1

[mrbitcoiner]

The day started with instabilities on my home server (x86_64 running Debian 12). Could not access it via SSH and had to login locally.

I've seen in my custom script that failures sending heartbeats to my monitoring database started at 03:28 UTC and stopped at 09:22 UTC, when I SIGKILLED lightningd.

After logging in, htop showed me that lightningd was using basically all my ram and swapfile.

After sending a SIGKILL to lightningd, my system become stable again.

Now, lightningd can not start again, as the logs show:

2025-02-20T09:20:51.543Z UNUSUAL plugin-bcli: bitcoin-cli: finished bitcoin-cli -rpcclienttimeout=3600 gettxout 77a42df1d5725249460feb1d643d9635df917a284752ee36d790cb76db882504 1 (10179 ms)
2025-02-20T09:21:16.049Z UNUSUAL plugin-bcli: bitcoin-cli: finished bitcoin-cli -rpcclienttimeout=3600 gettxout f775a5f7b41c0e714b53aefb0d187d7e4d54f30f5396c49f805719b1427151b0 4 (13443 ms)
2025-02-20T09:21:20.170Z UNUSUAL plugin-bcli: bitcoin-cli: finished bitcoin-cli -rpcclienttimeout=3600 estimatesmartfee 12 ECONOMICAL (14005 ms)
2025-02-20T09:35:57.375Z INFO    lightningd: v24.11.1
2025-02-20T09:35:57.572Z INFO    plugin-wss-proxy: Killing plugin: disabled itself: No python3 binary found
2025-02-20T09:35:57.573Z INFO    plugin-clnrest: Killing plugin: disabled itself: No python3 binary found
2025-02-20T09:35:57.573Z UNUSUAL plugin-bookkeeper: topic 'utxo_deposit' is not a known notification topic
2025-02-20T09:35:57.573Z UNUSUAL plugin-bookkeeper: topic 'utxo_spend' is not a known notification topic
2025-02-20T09:36:17.666Z INFO    plugin-bcli: bitcoin-cli initialized and connected to bitcoind.
2025-02-20T09:36:17.727Z UNUSUAL lightningd: Waiting for bitcoind to catch up (884553 blocks of 884562)
2025-02-20T09:37:27.427Z UNUSUAL plugin-bcli: bitcoin-cli: finished bitcoin-cli -rpcclienttimeout=3600 getblockhash 884530 (69695 ms)
2025-02-20T09:38:36.193Z UNUSUAL plugin-offers: Killing plugin: timed out before replying to init
2025-02-20T09:38:36.193Z **BROKEN** plugin-offers: Plugin marked as important, shutting down lightningd!

The latency in the calls to bitcoind is normal, as the bitcoind is running in a slow hard disk drive.

Any idea of what is going on?

I'm running on my own container image built from source

[mrbitcoiner]

I have other nodes running on v0.11.2 and never had any issues of that kind. This is the first week that I'm testing v24.11.1.

[mrbitcoiner]

I started the node again and it is working. Using memory limit of 1GB to the container, I will keep monitoring if the issue persists.