This repository has been archived by the owner on Nov 6, 2020. It is now read-only.
Fix performance issue importing Kovan blocks #9914
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
ngotchac
added
A0-pleasereview 🤓
andresilva
approved these changes
Nov 13, 2018
tomusdrw
added
A8-looksgood 🦄
tomusdrw
reviewed
Nov 13, 2018
| @@ -815,10 +816,11 @@ impl AuthorityRound { | |||
| None | |||
| } | |||
| }) | |||
| .while_some(); | |||
| .while_some() | |||
| .take_while(|&(h, _)| h != epoch_transition_hash); | |||
There was a problem hiding this comment.
So that limits the iterator, can you elaborate a little bit why it's valid to do so?
There was a problem hiding this comment.
I discussed this with @ngotchac. Aura only used to track finality for safely triggering validator set changes, therefore we were only interested in finalizing blocks within the current epoch, so this take_while was the original behavior. Recently we added finality tracking to the client so we updated Aura to track finality until the last finalized block. On chains where there isn't a majority of validators online the unfinalized subchain just keeps growing and this can take some time to build (although if the subchain is properly cached this cost is avoided).
|
#9692 is in beta, so should we backport this PR to beta as well? |
|
@ordian yes we should 👍 |
ordian
pushed a commit
that referenced
this pull request
Nov 14, 2018
Tbaut
added a commit
that referenced
this pull request
Nov 14, 2018
* Bump beta to version 2.2.1 * fix: Intermittent failing CI due to addr in use (#9885) Allow OS to set port at runtime * Use Weak reference in PubSubClient (#9886) * Fix json tracer overflow (#9873) * Fix json tracer overflow * Replace trace_executed with a direct trace push * Remove unused variable * Add test for 5a51 * Remove duplicate json! * Fix docker script (#9854) * Dockerfile: change source path of the newly added check_sync.sh (#9869) * Allow to seal work on latest block (#9876) * Allow to seal work on latest block. * Test from @todr to check sealing conditions. * gitlab-ci: make android release build succeed (#9743) * use docker cargo config file for android builds * make android build succeed * ethcore: use Machine::verify_transaction on parent block (#9900) * ethcore: use Machine::verify_transaction on parent block also fixes off-by-one activation of transaction permission contract * ethcore: clarify call to verify_transaction * foundation: #6692865, ropsten: #4417537, kovan: #9363457 * Remove rust-toolchain file (#9906) * EIP-712 implementation (#9631) * EIP-712 impl * added more tests * removed size parsing unwrap * corrected TYPE_REGEX to disallow zero sized fixed length arrays, replaced LinkedHashSet with IndexSet, added API spec to docs, fixed Type::Byte encoding branch * use Option<u64> instead of u64 for Type::Array::Length * replace `.iter()` with `.values()` Co-Authored-By: seunlanlege <seunlanlege@gmail.com> * tabify eip712.rs * use proper comments for docs * Cargo.lock: revert unrelated changes * tabify encode.rs * EIP 191 (#9701) * added sign_191 rpc method * fixed hash_structured_data return type * added ConfirmationPayload::SignMessage for non-prefixed signatures, added tests for sign191 * renamed WithValidator -> PresignedTransaction * rename applicationData to data in test * adds docs for EIP191Version, renamed SignRequest to EIP191SignRequest * light-fetch: Differentiate between out-of-gas/manual throw and use required gas from response on failure (#9824) * fix start_gas, handle OOG exceptions & NotEnoughGas * Change START_GAS: 50_000 -> 60_000 * When the `OutOfGas exception` is received then try to double the gas until it succeeds or block gas limit is reached * When `NotEnoughBasGas error` is received then use the required gas provided in the response * fix(light-fetch): ensure block_gas_limit is tried Try the `block_gas_limit` before regard the execution as an error * Update rpc/src/v1/helpers/light_fetch.rs Co-Authored-By: niklasad1 <niklasadolfsson1@gmail.com> * simplify cargo audit * Use block header for building finality (#9914) * ci: nuke the gitlab caches (#9855)
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
The
blockpassed toself.engine.ancestry_actionsdidn't have all the Header's data, thus theblock.header.hash()was wrong, and the conditionfinality_checker.subchain_head() != Some(*chain_head.parent_hash())inAuthorityRound::build_finalitywould never befalse, so we would check all ancestry for every imported blocks, until a finalised one was to be found. This was very costly in same cases, eg. when first syncing the chain: on currentmasterI would sync at 50 blocks/s, with this PR (as on stable) it goes at about 1_000 blocks/s.