[Prater] Late gossip blocks received

[twoeths]

Describe the bug

This is monitored in Prater nodes

This only happens in prater, I don't see this issue in mainnet and Ropsten

Expected behavior

No late block received

Steps to Reproduce

Run Prater

[dapplion]

@tuyennhv While investigating this issues is valuable this type of issues usually stay open in the Lodestar for a very long time. To organize this better I would suggest:

• Have some way to group this type of issues: one parent issue that links of this type of no-immediate-action-issues, and / or a label
• Try to gather more info locally, or spent some time investigating the issue
• Figure out what would be next steps to tackle the issue

If the purpose of this issue is only to journal odd behaviour we should have some way to check if this is Lodestar specific of from a network only. For this you could:

• Contrast with each prater node this metric. Are all nodes affected? Only this one? Why?
• Run other non-Lodestar nodes in Prater to contract findings

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 15 days if no further activity occurs. Thank you for your contributions.

[philknows]

Based on v1.1.0 metrics, this doesn't seem to be an issue anymore. We're seeing ~1 sec range now @tuyennhv . Feel free to close if agreed.

[twoeths]

It's still an issue in Goerli from what I see, this is from a stable node (5 first days with v1.0.0 and 2 last days with v1.1.0, they are the same)

It's better on mainnet nodes

[dapplion]

@tuyennhv Can you plot block times with an histogram like here? It will illustrate better the distribution https://twitter.com/dapplion/status/1575431858522886144

[twoeths]

not an issue anymore since v1.2.0, below is 10m rate interval on the stable 1k node: