[Bug] Bluebox disseminating same block to compact to multiple machines

[woodycal]

What happened?

As per title when running a bluebox-timelord in cluster mode, bluebox disseminates same blocks to compact to multiple machines, I have found one proof of this which i added in log output.

I have target_uncompact_proofs: 100 and send_uncompact_interval 300 which is more then enough work because i am averaging 30 proofs a hour with 3 machines currently due to high compute(asic timelords).

Version

2.21

What platform are you using?

Linux

What ui mode are you using?

CLI

Relevant log output

**Main timelord bluebox log**
2024-03-24T17:58:10.785 full_node chia.full_node.full_node: INFO     Replaced compact proof at height 3911738

2024-03-24T17:58:48.411 full_node chia.full_node.full_node: INFO     Replaced compact proof at height 3911738
2024-03-24T17:58:48.466 TLauncher chia.timelord.timelord_launcher: INFO     VDF client 0: VDF Client: Discriminant = -0xfd40a0c0d09e323d695321ac2dddda68fa706722ba784e7ba80ccd031060b25c6417944f27e1f140e7bb016655e31c2ef324d7243f228e40b8a1>
Got simple weso proof: 0200650b79a0060c29fad654fc723c049cce7b4c98abba486606d570b69abcfb75d5c75986f34a07d9c8712c56379cdd55c37f245b8d07caac79085f53349d2e6708d6a2471a9ff6e100991f3b3e27270c2b5f20c23e9d9af42422b394593e920c0208030000e9478f29e>
VDF Client: Sending proof
VDF Client: Sent proof
VDF loop finished. Total iters: 172389354
VDF Client: Stopped everything! Ready for the next challenge.

**And log from a compute bluebox**
2024-03-24T17:58:10.867 TLauncher chia.timelord.timelord_launcher: INFO     VDF client 3: VDF Client: Discriminant = -0xe5823ccd7e615936cb1d206b9fe8f57a6e6cbbf82ab2ff5c9483194ca6c465ff39278b5e31bdbf38df94b8b0362d29c4abe49266d450ce21311f>
Got simple weso proof: 02009fa843bad42550f31c1d9f152435613f77499869f4bdddb611af3993452231069296cc7f3c1587e88ddcc713963f8c1ef0619919d7f60182a73df55ee838eb629c07ea2d50401d73b81dd3689589292250db4e63411531b990a3104be8b071630100020000c731703>
VDF Client: Sending proof
VDF Client: Sent proof
VDF loop finished. Total iters: 140058306

[wjblanke]

Ya I bet it just broadcasts it out. Let us see if there is a way to make it work better when you are running a cluster.

Thanks for being a blueboxer!

[woodycal]

2024-03-27T17:10:05.532 full_node chia.full_node.full_node: INFO Replaced compact proof at height 3562487
2024-03-27T17:10:08.341 full_node chia.full_node.full_node: INFO Replaced compact proof at height 3562487

Found another one from a quick look through so it must be uncommon.

Just to put this into context my main node is not overly powerful and my main machines are windows so i run virtual box ubuntu compute timelord from windows as and when. It saves time and space then running separate nodes and having to wait for syncing.

I don't really have threadripper or other really high core count cpu but i am willing to bet it could do the same with them.

pre asic timelord this might been a non issue tbh but understand if you leave this if its too much work.