markqvist
LXMD running a propagation node stops working after some time
Describe the Bug I have LXMD set up to run a propagation node using a systemd service and the following settings:
- autopeer = yes
- autopeer_maxdepth = 10
- propagation_transfer_max_accepted_size = 32768
- message_storage_limit = 100000
- max_peers = 100
After running for ~1 day I can no longer interact with the node. Other services running with the same shared reticulum instance are fine.
Trying to sync messages results in this error message Link establishment failed
To Reproduce Run LXMD for some time with high peer and storage limit? I have no other idea on how to reproduce this.
Expected Behavior Being able to establish a link and sync messages.
Logs & Screenshots
The last logs collected by systemd (lxmd -p; loglevel 7; the last log entry is from ca. 20 minutes ago):
[2025-08-03 12:02:58] [Debug] No path to peer <38285e0023b4ab9b08faa506c2cec1c9> exists, requesting...
[2025-08-03 12:03:06] [Debug] Path request was not answered, retrying sync with peer <38285e0023b4ab9b08fa>
[2025-08-03 12:03:54] [Debug] Calculating peer distribution queue mappings...
[2025-08-03 12:03:54] [Debug] Distribution queue mapping completed in 0s
[2025-08-03 12:04:03] [Debug] Selecting peer to sync from 63 waiting peers.
[2025-08-03 12:04:03] [Debug] Selected waiting peer 0: <43e8e386d6e5d3fe1cafe3faf2d0a1a9>
[2025-08-03 12:04:03] [Debug] Initiating LXMF Propagation Node sync with peer <43e8e386d6e5d3fe1cafe3faf2d>
[2025-08-03 12:04:03] [Debug] No path to peer <43e8e386d6e5d3fe1cafe3faf2d0a1a9> exists, requesting...
[2025-08-03 12:04:10] [Debug] Path request was not answered, retrying sync with peer <43e8e386d6e5d3fe1caf>
[2025-08-03 12:04:58] [Debug] Calculating peer distribution queue mappings...
Running lxmd --status:
[2025-08-03 12:25:02] [Error] Getting lxmd statistics timed out, exiting now
Logs from nomadnet trying to sync with the node:
[Debug] No path known for message download from propagation node <f141f039b3b88b7a2d5c6048c7adaafb>. Requesting path...
[Extra] Valid announce for <f141f039b3b88b7a2d5c6048c7adaafb> 15 hops away, received via <beda0b851cab83c88c246792c261175e> on BackboneInterface[aspark TCP/aspark.uber.space:44860]
[Debug] Destination <f141f039b3b88b7a2d5c6048c7adaafb> is now 15 hops away via <beda0b851cab83c88c246792c261175e> on BackboneInterface[aspark TCP/aspark.uber.space:44860]
[Debug] Establishing link to <f141f039b3b88b7a2d5c6048c7adaafb> for message download
[Debug] Signalling link MTU of 16.38 KB for link
[Debug] Establishing link with mode AES_256_CBC
[Extra] Registering link <2ecdfaf0b20b100d6f5c3b6aedbec62d>
[Debug] Link request <2ecdfaf0b20b100d6f5c3b6aedbec62d> sent to <lxmf.propagation.68051716d078ed8565b8748e6c77ee10:f141f039b3b88b7a2d5c6048c7adaafb>
[Extra] Establishment timeout is 1m and 36.0s for link request <2ecdfaf0b20b100d6f5c3b6aedbec62d>
...
[Verbose] Link establishment timed out
[Debug] Trying to rediscover path for <f141f039b3b88b7a2d5c6048c7adaafb> since an attempted link was never established
[Debug] Path to <f141f039b3b88b7a2d5c6048c7adaafb> timed out and was removed
[Extra] Removed 1 path
[Extra] Valid announce for <f141f039b3b88b7a2d5c6048c7adaafb> 15 hops away, received via <beda0b851cab83c88c246792c261175e> on BackboneInterface[aspark TCP/aspark.uber.space:44860]
[Debug] Destination <f141f039b3b88b7a2d5c6048c7adaafb> is now 15 hops away via <beda0b851cab83c88c246792c261175e> on BackboneInterface[aspark TCP/aspark.uber.space:44860]
[Debug] Cleaned outbound propagation link
From what I can tell, it fails to establish a link to my node. Syncing with other nodes works fine.
System Information
- OS and version: Raspberry Pi OS based on debian 12 (up to date)
- Python version: 3.11.2
- Program version: lxmd 0.8.0, rnsd 1.0.0
Same problem here. The only thing I can say, is that the storage limit does not seem to have an effect. I tried the default of 500 megabytes and also 2000 megabytes. The problem remains the same.
Unfortunately I forgot at what version exactly the problem started, but it also occurred in version 0.7.1.
Configuration
- autopeer = yes
- autopeer_maxdepth = 4
- propagation_transfer_max_accepted_size = 256
- message_storage_limit = 2000
- max_peers = 50
System Information
- OS and version: Debian 12 x86_64 (up to date)
- Python version: 3.11.2
- Program version: lxmd 0.8.0, rnsd 1.0.0
Wow, 50 and 100 max peers? That seems like a lot. Are you attached to the testnet? I see 63 peers in the first posts logs...
The testnet seems to be in a somewhat unstable state due to the 128/256 key size encryption thing. But I would hope that those peers would be dropped as they cannot be communicated with. But I think that is untested.
Yes, the Propagation Node is attached to the testnet. It only uses default settings. max_peers = 50 is the default as well.
Maybe you are right and the problem had something to do with the encryption key size change. The funny thing is, I didn't have a problem since complaining about it here in August.