-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG] Delayed message propagation on network setup #6396
Comments
@travisperson Was this much shorter setup time (tipset 42) close to normal for when you've done this in the past? Or is there still a noticeable difference? |
Generally the additional miners will be setup within blocks 11-20. The genesis miner will initialize in ~10 tipsets, and the rest of the miners will come in just a bit after. So 42 is longer than normal, but not by a lot. For example here is the beginning of the calibration network:
|
Next step is to investigate butterfly nodes that are manifesting this issue. |
Update to:
The butterfly nodes were running into a Bellperson deadlock, which has since been fixed. That may have been the rootcause here, but it's unlikely -- the nodes were manifesting different symptoms. However, we have since launched multiple networks and aren't seeing reports of this issue. If the performance metrics added in #6453 don't show anything concerning, I'd be comfortable closing this and moving forward. |
Next step is for @magik6k to look at performance metrics on his node. |
On the first reset of the calibration network it doesn't appear this issue occurred at all. Even got some messages into blocks during the initial setup phase.
|
Closing because this seems to have been a mysterious occurrence. |
Ran into this issue when deploying a network based on
feat/nv13-1.11
(40cc29d)During a network setup I observed a prolonged initialization of the non-genesis miners (t01001, t01002) after the genesis miner (t01000) was up and running. After some inspection I found that the initialization message for t01001 and t01002 were not showing up in t01000 mempool, or each others mempools. The messages were only being propagated to the bootstrap peers on the network and no other peers.
All nodes were fully connected, all peer scores were non-negative, and all nodes were fully synced with miner t01000.
Shortly after message started to propagate and t01002 mine its first block.
I reset the network after this, and observed a much shorter setup time, with t01001 and t01002 joining at tipset height 42.
The text was updated successfully, but these errors were encountered: