Display block and chunk producers of an epoch #7222
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
There are many lovable aspects to `setup_mock_all_validators`! What this PR deals with is "tying the knot" infra. `setup_mock_validator` is based on creating `n` validators and setting up the mock network between them. To create a mock network, you need all `n` validators, and to create a validator, you need a mock network. The cycle is broken at runtime, in three different ways: * inside `setup_mock_all_validators`, we have `locked_connectors` * at some call-sites, we overwrite the whole `peer_manager_mock` * yet at other call-sites, we essentially create a copy of `locked_connectors` This PR unifies that to a single infra: * there's a single `Arc<OnceCell<Vec<Addr<Client>>>>` inside `setup_mock_all_validators` * `OnceCell::wait` is used as a barrier * The `&[Addr<CLient>]` are passed in as an argument to `peer_manager_mock`, so that it doesn't have to re-invent a way to get a hold of those This changes the signature of the function, so the diff is pretty large. It is designed to be reviewed per-commit.
The condition to check for high variance of measurements so far has been that no single sample is further than 15% from the mean away. We often missed that when running with 5 iterations and one warm-up iteration. The counter-intuitive side of this check is that is becomes less likely that variance is considered low enough, if more iterations are added. The new condition computes actual variance and standard deviation. Now we expect that standard deviation is lower than 10%. Also add an outer loop to block overhead computation. Usually there is a loop + a number of actions inside a block. But this estimation has empty blocks and therefore each measurement so far was just a single apply block call. Grouping 10 empty blocks together for each measurement and adding an outer loop for `--iters` reduces the overall variance.
Got this one spuriously(?) firing in some test with nightly features. Let's make the failure a bit more helpful here!
trie_key module stores separators and keys as slices even though they are all just one-byte long. There are parts of the code which assume the length to be one and won’t work if the length ever changes. Simplify the code by keeping the values as plain bytes rather than slices so that it’s explicit in their type that they are always just one byte.
* eagarly compute assignemt of validators to blocks & chunks, which I find a bit easier to wrap my head around (you can print `EpochValidatorSet`s and just see the result) * compress the code by centralizing cfg handling
Closes #7203 There are a few structs that do not need to derive `Serialize` and `Deserialize`. Remove the derives highlights that there are some structs that we do not need at all and that we do not need to define custom deserializer: `bytes_as_string`. Tests ------- - `cargo check --tests --benches` does not complain. - Other existing tests
Explicitly tracking the metric inside `GasCost` is awkward for zero values and it requires the estimator configuration to be passed around more than strictly necessary. This change makes the metric implicit, by virtue of having or not having the measurement values stored inside. For conversion to gas, and when printing, `ICount` measurements are given priority over `Time` measurements when for both values exist. This could be changed, if both need to be tracked simultaneously, for now only one of them is of interest at the time.
Creating a function to get int rocks db properties for every column and adding a way to export it as a prometheus metric. Reorganising parse_statistics to write stats into provided result param.
mina86
approved these changes
Jul 20, 2022
mm-near
approved these changes
Jul 21, 2022
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.
No description provided.