-
Notifications
You must be signed in to change notification settings - Fork 111
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
Handle duplicate block errors #1372
Comments
We should wait to examine this until after implementing block gossip (#1082, #1328), since the sync component is only supposed to be responsible for initial block sync, not ongoing gossip-based sync. So the sync component exhausting tips and restarting isn't in itself a problem. But this behavior doesn't sound good, and should be investigated:
|
I'm not seeing these errors on Mainnet any more. We should re-test to make sure they're not happening. |
With current
|
I'm seeing these errors a long way from the tip on testnet. |
We might want to remove the spantrace - it makes the error seem much more serious than it actually is. |
I can look into hiding it in the logs but I don't think we should get rid of it. I'm thinking maybe just don't show it in logs by default but always include it in the generated issue url. Also it might be useful to drop the level on this span so that it gets included in the spantrace #[instrument(level = "debug", skip(self), fields(%hash))]
pub async fn download_and_verify(&mut self, hash: block::Hash) -> Result<(), Report> {
if self.cancel_handles.contains_key(&hash) {
return Err(eyre!("duplicate hash queued for download"));
} |
We've told users to ignore these errors in the README, so I disabled the issue URL in #1517
I think an info-level span would be too verbose. Instead, I included the hash in the error message in #1517 |
This is fixed or covered by #862 |
Version
zebrad
main with all outstanding workaround PRs applied, including #1370 and #1371.Description
As Zebra sync approaches the tip, it starts encountering duplicate block errors:
Once these errors pass, the syncer gets into a loop where it has exhausted prospective tips. It seems to lag 10-100 blocks behind the current chain tip. It's unclear why short block locator requests and responses don't work.
Suggested Solutions
Detailed Logs
When the Zcash Mainnet tip was around 1053180:
When the Zcash Mainnet tip was around 1053200:
The text was updated successfully, but these errors were encountered: