-
Notifications
You must be signed in to change notification settings - Fork 1.7k
Description
System information
Geth version: geth version 1.5.17-d689f3bb-20250624
(geth 1.5.16 has the same problems)
OS & Version: Linux
After upgrading to v1.5.16 and later v1.5.17, the node is frequently unable to discover peers or maintain stable connections. Syncing becomes extremely unreliable — blocks are downloaded sporadically every few minutes, and then the node resumes searching for peers with almost no success.
This appears to be the same issue as described in #3168, which was unfortunately closed without resolution, though the underlying problem still persists.
Typical log output:
Jul 07 09:35:08 bsc1 sh[3540979]: INFO [07-07|09:35:08.731] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:35:18 bsc1 sh[3540979]: INFO [07-07|09:35:18.884] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:35:29 bsc1 sh[3540979]: INFO [07-07|09:35:29.333] Looking for peers peercount=1 tried=44 static=57
........
Jul 07 09:36:28 bsc sh[3540979]: INFO [07-07|09:36:28.885] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:36:39 bsc sh[3540979]: INFO [07-07|09:36:39.334] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:36:53 bsc sh[3540979]: INFO [07-07|09:36:53.732] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:37:03 bsc sh[3540979]: INFO [07-07|09:37:03.886] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:37:14 bsc sh[3540979]: INFO [07-07|09:37:14.334] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:37:28 bsc sh[3540979]: INFO [07-07|09:37:28.731] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:37:38 bsc sh[3540979]: INFO [07-07|09:37:38.887] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:37:49 bsc sh[3540979]: INFO [07-07|09:37:49.335] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:38:03 bsc sh[3540979]: INFO [07-07|09:38:03.733] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:38:13 bsc sh[3540979]: INFO [07-07|09:38:13.886] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:38:24 bsc sh[3540979]: INFO [07-07|09:38:24.336] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:38:38 bsc sh[3540979]: INFO [07-07|09:38:38.734] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:38:48 bsc sh[3540979]: INFO [07-07|09:38:48.888] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:38:59 bsc sh[3540979]: INFO [07-07|09:38:59.337] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:39:13 bsc sh[3540979]: INFO [07-07|09:39:13.735] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:39:23 bsc sh[3540979]: INFO [07-07|09:39:23.888] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:39:34 bsc sh[3540979]: INFO [07-07|09:39:34.345] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:39:48 bsc sh[3540979]: INFO [07-07|09:39:48.735] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:39:58 bsc sh[3540979]: INFO [07-07|09:39:58.889] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:40:09 bsc sh[3540979]: INFO [07-07|09:40:09.346] Looking for peers peercount=1 tried=44 static=57
Jul 07 09:40:23 bsc sh[3540979]: INFO [07-07|09:40:23.736] Looking for peers peercount=1 tried=5 static=57
Jul 07 09:40:33 bsc sh[3540979]: INFO [07-07|09:40:33.889] Looking for peers peercount=1 tried=7 static=57
Jul 07 09:40:44 bsc sh[3540979]: INFO [07-07|09:40:44.346] Looking for peers peercount=1 tried=44 static=57
All suggestions from issue #3168 and additional docs have been implemented:
- Latest known BootNodes added (from the another issue)
- Regenerated nodekey to ensure new enode
Not Caused by Hardware or Hosting
There are no hardware limitations: CPU, memory, disk, and network bandwidth are all well within safe operating limits.
This is not a hosting issue: I run multiple ethlike nodes (e.g., Ethereum, Polygon, Arbitrum) on the same infrastructure without any issues — only BSC nodes are affected.
Impact
After hardfork and mandatory upgrade, the current node version is effectively unusable for stable block syncing. Many operators are reporting the same issue but are left without workaround due to closure of the previous ticket.