NeoFS BlockFetcher hangs on shut down #1510

New issue

Open

opened 2025-12-28 17:16:41 +00:00 by sami · 0 comments

sami commented

2025-12-28 17:16:41 +00:00

Owner

Originally created by @AnnaShaleva on GitHub (May 3, 2025).

Current Behavior

Once all blocks are fetched from NeoFS, BlockFetcher service hangs on shutdown:

2025-05-03T09:59:09.815+0300	INFO	runtime log	{"tx": "727a67d07bfcacaf79e6f26c18ce9ed58cfd43e89bb2bba1dc649a0a3ce4d2d4", "script": "c4576ea5c3081dd765a17aaaa73d9352e74bdc28", "msg": "update state of the network map candidate"}
2025-05-03T09:59:09.817+0300	INFO	runtime log	{"tx": "99a602f157b11f156aaeee04e964ebffebfd5238451a2a7f917064be8374f198", "script": "c4576ea5c3081dd765a17aaaa73d9352e74bdc28", "msg": "update state of the network map candidate"}
2025-05-03T09:59:10.347+0300	INFO	persisted to disk	{"blocks": 443, "keys": 6502, "headerHeight": 14406027, "blockHeight": 14406027, "took": "34.592594ms"}
2025-05-03T09:59:44.247+0300	INFO	NeoFS BlockFetcher service: no block found with index 14409585, stopping
2025-05-03T09:59:44.247+0300	INFO	shutting down NeoFS BlockFetcher service	{"force": false}

Expected Behavior

BlockFetcher should be shut down normally once all blocks are fetched, P2P sync should start.

Possible Solution

Debug the shutdown routine.

Context

I use SkipIndexFilesSearch: true mode, but it looks like index files enable mode is also hanging sometimes. Also, the node doesn't have any peers (I'm mentioning it just in case if it might be important). NeoFSStateSyncExtensions are off, so it's a standard full archival node sync. I also use DefaultSearchBatchSize = 4000 which is not the default setting for the node.

Your Environment

Version of the product used: NeoGo 0.109.0

Originally created by @AnnaShaleva on GitHub (May 3, 2025). ## Current Behavior Once all blocks are fetched from NeoFS, BlockFetcher service hangs on shutdown: ``` 2025-05-03T09:59:09.815+0300 INFO runtime log {"tx": "727a67d07bfcacaf79e6f26c18ce9ed58cfd43e89bb2bba1dc649a0a3ce4d2d4", "script": "c4576ea5c3081dd765a17aaaa73d9352e74bdc28", "msg": "update state of the network map candidate"} 2025-05-03T09:59:09.817+0300 INFO runtime log {"tx": "99a602f157b11f156aaeee04e964ebffebfd5238451a2a7f917064be8374f198", "script": "c4576ea5c3081dd765a17aaaa73d9352e74bdc28", "msg": "update state of the network map candidate"} 2025-05-03T09:59:10.347+0300 INFO persisted to disk {"blocks": 443, "keys": 6502, "headerHeight": 14406027, "blockHeight": 14406027, "took": "34.592594ms"} 2025-05-03T09:59:44.247+0300 INFO NeoFS BlockFetcher service: no block found with index 14409585, stopping 2025-05-03T09:59:44.247+0300 INFO shutting down NeoFS BlockFetcher service {"force": false} ``` ## Expected Behavior BlockFetcher should be shut down normally once all blocks are fetched, P2P sync should start. ## Possible Solution Debug the shutdown routine. ## Context I use `SkipIndexFilesSearch: true` mode, but it looks like index files enable mode is also hanging sometimes. Also, the node doesn't have any peers (I'm mentioning it just in case if it might be important). `NeoFSStateSyncExtensions` are off, so it's a standard full archival node sync. I also use `DefaultSearchBatchSize = 4000` which is not the default setting for the node. ## Your Environment * Version of the product used: NeoGo 0.109.0