Increase block validation timeouts

[jvff]

Motivation

The full synchronization test currently times out after running for 6 hours. One possible cause is that during synchronization the block validation may time out a few times, which leads to a few minutes of delays.

Solution

Increase the UTXO lookup timeout and the block validation time out to see if the synchronization finishes earlier.

Review

Reviewer Checklist

• Code implements Specs and Designs
• Tests for Expected Behaviour
• Tests for Errors

Follow Up Work

[teor2345]

This could be a bug in the recently merged PR #4149, or an intermittent error:

WARNING: Default device-name for disk name [zebrad-cache-28dc985-mainnet-checkpoint] will be [zebrad-cache-28dc985-mainnet-checkpoint] because it is being mounted to a container with [--container-mount-disk]
ERROR: (gcloud.compute.instances.create-with-container) Could not fetch resource:

• The resource 'projects/zealous-zebra/global/images/zebrad-cache-main-e5f00c5-v23-mainnet-checkpoint' is not ready

https://github.com/ZcashFoundation/zebra/runs/6103901143?check_suite_focus=true#step:7:95

Maybe @gustavovalverde can help diagnose?

[jvff]

This did not lead to a performance increase that's enough to solve the issue (#4155).

This PR can be closed, but it might be useful for reference because it contains the workaround I used when synchronization gets stuck on a single block. This can happen on debug builds or on release builds using trace-level logging. It could also happen on some lower-end computers.

[teor2345]

This did not lead to a performance increase that's enough to solve the issue (#4155).

This PR can be closed, but it might be useful for reference because it contains the workaround I used when synchronization gets stuck on a single block. This can happen on debug builds or on release builds using trace-level logging. It could also happen on some lower-end computers.

I'd like to fix this bug for low-end computers, if we can do it without breaking the full sync test.

Can we leave this PR in draft, fix the sync speed, and then re-test to make sure it doesn't make the sync worse?

[teor2345]

@Mergifyio update

[mergify]

update

✅ Branch has been successfully updated

[teor2345]

I am updating this branch to delete the clean action, which causes other actions to fail.

[teor2345]

@Mergifyio update

[mergify]

update

✅ Branch has been successfully updated

[teor2345]

I'm running a full sync test here:
https://github.com/ZcashFoundation/zebra/actions/runs/2267417739

If it takes around 5h50m, we should merge this PR, because it fixes part of the bug in PR #4286.

[oxarbitrage]

So the full sync tests was skipped, can we relaunch this to check how much time it takes ?

[teor2345]

So the full sync tests was skipped, can we relaunch this to check how much time it takes ?

The test I ran manually took 5h47m, so it is about the same speed.

[teor2345]

Required to fix bugs like #4286.