Prepare the GraphQL service for the switching to async methods
#2334
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
netrome
previously approved these changes
Oct 12, 2024
xgreenx
force-pushed
the
feature/prepare-graphql-for-async
branch
from
7cb4231 to
ab5e940
Compare
rymnc
approved these changes
Oct 14, 2024
netrome
approved these changes
Oct 14, 2024
Merged
xgreenx
added a commit
that referenced
this pull request
Oct 14, 2024
## Version v0.40.0 ### Added - [2347](#2347): Add GraphQL complexity histogram to metrics. - [2350](#2350): Added a new CLI flag `graphql-number-of-threads` to limit the number of threads used by the GraphQL service. The default value is `2`, `0` enables the old behavior. - [2335](#2335): Added CLI arguments for configuring GraphQL query costs. ### Fixed - [2345](#2345): In PoA increase priority of block creation timer trigger compare to txpool event management ### Changed - [2334](#2334): Prepare the GraphQL service for the switching to `async` methods. - [2310](#2310): New metrics: "The gas prices used in a block" (`importer_gas_price_for_block`), "The total gas used in a block" (`importer_gas_per_block`), "The total fee (gwei) paid by transactions in a block" (`importer_fee_per_block_gwei`), "The total number of transactions in a block" (`importer_transactions_per_block`), P2P metrics for swarm and protocol. - [2340](#2340): Avoid long heavy tasks in the GraphQL service by splitting work into batches. - [2341](#2341): Updated all pagination queries to work with the async stream instead of the sync iterator. - [2350](#2350): Limited the number of threads used by the GraphQL service. #### Breaking - [2310](#2310): The `metrics` command-line parameter has been replaced with `disable-metrics`. Metrics are now enabled by default, with the option to disable them entirely or on a per-module basis. - [2341](#2341): The maximum number of processed coins from the `coins_to_spend` query is limited to `max_inputs`. ## What's Changed * fix(gas_price_service): service name and unused trait impl by @rymnc in #2317 * Do not require build of docker images to pass CI by @xgreenx in #2342 * Prepare the GraphQL service for the switching to `async` methods by @xgreenx in #2334 * Limited the number of threads used by the GraphQL service by @xgreenx in #2350 * Increase priority of timer over txpool event by @xgreenx in #2345 * Disable flaky `test_poa_multiple_producers` test by @rafal-ch in #2353 * feat: CLI arguments for configuring GraphQL query costs. by @netrome in #2335 * Add graphql query complexity histogram metric by @AurelienFT in #2349 * Updated all pagination queries to work with the `Stream` instead of `Iterator` by @xgreenx in #2341 * Avoid long heavy tasks in the GraphQL service by @xgreenx in #2340 * Add more metrics by @rafal-ch in #2310 **Full Changelog**: v0.39.0...v0.40.0 --------- Co-authored-by: Rafał Chabowski <rafal.chabowski@fuel.sh> Co-authored-by: acerone85 <andrea.cerone@gmail.com> Co-authored-by: rymnc <43716372+rymnc@users.noreply.github.com> Co-authored-by: Rafał Chabowski <88321181+rafal-ch@users.noreply.github.com>
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.
To avoid blocking all threads with heavy sync requests to the database, we decided to move the main database lookups to a separate thread pool.
After we introduced
ReadViewin the GraphQL service, we duplicated all ports for this type to avoid breaking many places in the codebase(plus simplifying the review process of the change that caused the creation ofReadView). But after adding theReadViewtype, we don't need ports on this type anymore since it is an internal database wrapper type of the GraphQL service, and it has full control over how it works. So, it was an old tech debt to remove the usage of ports. The time is come. To simplify the introduction of theasyncand not affect ports with it, this PR decouples ports andReadView.This PR prepares the switching from
synctoasynclogic(it will be a follow-up PR).Before requesting review