dynamic block time modifications #983
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
It looks like the volatility is worse when one reduces the target time down to 2 seconds EDIT: Maybe I just got lucky the first run. It seems the standard deviation is around 0.5s |
evan-forbes
approved these changes
Mar 20, 2023
Comment on lines
-674
to
+671
| elapsed := cs.CommitTime.Sub(cs.StartTime) | ||
| nst := cs.config.TargetHeightDuration - elapsed | ||
| if nst < time.Millisecond*1 || elapsed < 0 { | ||
| nst = time.Millisecond * 1 | ||
| } | ||
| cs.eventCollector.WritePoint("consensus", map[string]interface{}{ | ||
| "elapsed_info": []interface{}{cs.Height, cs.Round, elapsed, nst}, | ||
| }) | ||
| cs.StartTime = cs.CommitTime.Add(nst) | ||
| cs.StartTime = cs.config.NextStartTime(cs.StartTime) |
There was a problem hiding this comment.
thanks for refactoring this back, I was meaning to move the logic back to NextStartTime
Comment on lines
-970
to
+976
| TimeoutPropose: 3000 * time.Millisecond, | ||
| TimeoutPropose: 2000 * time.Millisecond, | ||
| TimeoutProposeDelta: 500 * time.Millisecond, | ||
| TimeoutPrevote: 1000 * time.Millisecond, | ||
| TimeoutPrevoteDelta: 500 * time.Millisecond, | ||
| TimeoutPrecommit: 1000 * time.Millisecond, | ||
| TimeoutPrecommitDelta: 500 * time.Millisecond, | ||
| TargetHeightDuration: 3500 * time.Millisecond, | ||
| TargetHeightDuration: 3000 * time.Millisecond, |
There was a problem hiding this comment.
are these needed for the e2e test to pass or just arbitrary?
There was a problem hiding this comment.
I just wanted to make it smaller / quicker. Have not strong feelings on what these are set to
|
when increasing the target time, the volatility seems to decrease! I'm seeing ~.2s in variation for 6s target time, but could just be my machine |
Yeah I think we need to look over a longer range of time (and ideally with 100 nodes and a semi-decent load to see if those factors make an influence) |
evan-forbes
added a commit
that referenced
this pull request
Apr 20, 2023
…#965) * feat!: consider time already elapsed when waiting after the commit * chore: minor doc changes * chore: try a different default config time * chore: try increasing the config again * fix: use appropriate default time * fix: docs * chore: simplify addition and rename * chore: revert pointless go mod tidy change * chore: consistent config * chore: replace config comments * chore: fix a few remaining round -> height name changes * fix: lingering compile errors from merge * fix: silly bug * fix: use correct next start time * dynamic block time modifications (#983) * chore: remove event collector * Update test/maverick/consensus/state.go Co-authored-by: Callum Waters <[email protected]> * Update test/maverick/consensus/state.go Co-authored-by: Callum Waters <[email protected]> * chore: formatting --------- Co-authored-by: Callum Waters <[email protected]>
cmwaters
pushed a commit
that referenced
this pull request
Sep 20, 2023
* Add `CMT_HOME` (or remove it?) (#983) Closes #982 Added `CMT_HOME` everywhere `CMTHOME` is used. ### Notes to reviewers This could be fixed the opposite way, by removing the only reference to `CMT_HOME` in the code, and also the reference in `UPGRADING.md` (two lines of code). However, the reference in `UPGRADING.md`, which is part of our documentation, is already present in `v0.34.x` (not in `v0.37.x` though!). That's why this PR introduces `CMT_HOME` to work in equal conditions as `CMTHOME`. If reviewers lean toward removing `CMT_HOME` from the doc in `v0.34.x` (and unreleased `v0.38.x` and `main`), I can do it easily. --- #### PR checklist - [x] Tests written/updated - [ ] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [ ] Updated relevant documentation (`docs/` or `spec/`) and code comments (cherry picked from commit b7be568) # Conflicts: # cmd/cometbft/commands/root_test.go * Revert "Add `CMT_HOME` (or remove it?) (#983)" * Add `CMT_HOME` (or remove it?) (#983) Closes #982 Added `CMT_HOME` everywhere `CMTHOME` is used. This could be fixed the opposite way, by removing the only reference to `CMT_HOME` in the code, and also the reference in `UPGRADING.md` (two lines of code). However, the reference in `UPGRADING.md`, which is part of our documentation, is already present in `v0.34.x` (not in `v0.37.x` though!). That's why this PR introduces `CMT_HOME` to work in equal conditions as `CMTHOME`. If reviewers lean toward removing `CMT_HOME` from the doc in `v0.34.x` (and unreleased `v0.38.x` and `main`), I can do it easily. --- - [x] Tests written/updated - [ ] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [ ] Updated relevant documentation (`docs/` or `spec/`) and code comments --------- Co-authored-by: Sergio Mena <[email protected]>
cmwaters
added a commit
that referenced
this pull request
Jun 7, 2024
…#965) * feat!: consider time already elapsed when waiting after the commit * chore: minor doc changes * chore: try a different default config time * chore: try increasing the config again * fix: use appropriate default time * fix: docs * chore: simplify addition and rename * chore: revert pointless go mod tidy change * chore: consistent config * chore: replace config comments * chore: fix a few remaining round -> height name changes * fix: lingering compile errors from merge * fix: silly bug * fix: use correct next start time * dynamic block time modifications (#983) * chore: remove event collector * Update test/maverick/consensus/state.go Co-authored-by: Callum Waters <[email protected]> * Update test/maverick/consensus/state.go Co-authored-by: Callum Waters <[email protected]> * chore: formatting --------- Co-authored-by: Callum Waters <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Some data from a 4 validator network under transaction load with a 5 second target height:
Would be good to test this under "realistic" network environments i.e. most likely with testground.
PR checklist
.changelog(we useunclog to manage our changelog)
docs/orspec/) and code comments