Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

perf(db): Try yet another storage log pruning approach #2268

Merged
merged 1 commit into from
Jun 19, 2024
Merged

perf(db): Try yet another storage log pruning approach #2268

merged 1 commit into from
Jun 19, 2024

Conversation

slowli
Copy link
Contributor

@slowli slowli commented Jun 19, 2024 •
edited
Loading

What ❔

Structures storage log pruning differently by first loading primary keys for the latest logs in the pruned block range, and then range-removing older logs based on these PKs. Both of these queries are designed to use particular indexes, making them have predictable performance.

Why ❔

The current DB queries for storage log pruning sometimes use unpredictable indexes and have suboptimal performance.

Checklist

  • PR title corresponds to the body of PR (we generate changelog entries from PRs).
  • Tests for the changes have been added / updated.
  • Documentation comments have been added / updated.
  • Code has been formatted via zk fmt and zk lint.
  • Spellcheck has been run via zk spellcheck.

@slowli slowli marked this pull request as ready for review June 19, 2024 07:57
@slowli slowli added this pull request to the merge queue Jun 19, 2024
Merged via the queue into main with commit 3ee34be Jun 19, 2024
49 of 50 checks passed
@slowli slowli deleted the aov-pla-971-try-more-efficient-pruning-query-take2 branch June 19, 2024 11:05
@zksync-era-bot zksync-era-bot mentioned this pull request Jun 19, 2024
Merged
@slowli slowli mentioned this pull request Jun 19, 2024
Merged
4 tasks
github-merge-queue bot pushed a commit that referenced this pull request Jun 19, 2024
@slowli
refactor(db): Combine storage log pruning into single query (#2279)
7f4e6ac 
## What ❔

Minor follow-up for #2268
that combines both parts of log pruning into a single query.

## Why ❔

Easier to maintain and could be slightly more efficient since
intermediate data doesn't need to travel from Postgres to the node and
back.

## Checklist

- [x] PR title corresponds to the body of PR (we generate changelog
entries from PRs).
- [x] Documentation comments have been added / updated.
- [x] Code has been formatted via `zk fmt` and `zk lint`.
- [x] Spellcheck has been run via `zk spellcheck`.
gabrieldemian pushed a commit to vianetwork/via-server that referenced this pull request Jun 21, 2024
@slowli @gabrieldemian
perf(db): Try yet another storage log pruning approach (matter-labs#2268
7c0a9fd 
)

## What ❔

Structures storage log pruning differently by first loading primary keys
for the latest logs in the pruned block range, and then range-removing
older logs based on these PKs. Both of these queries are designed to use
particular indexes, making them have predictable performance.

## Why ❔

The current DB queries for storage log pruning sometimes use
unpredictable indexes and have suboptimal performance.

## Checklist

- [x] PR title corresponds to the body of PR (we generate changelog
entries from PRs).
- [x] Tests for the changes have been added / updated.
- [x] Documentation comments have been added / updated.
- [x] Code has been formatted via `zk fmt` and `zk lint`.
- [x] Spellcheck has been run via `zk spellcheck`.
gabrieldemian pushed a commit to vianetwork/via-server that referenced this pull request Jun 21, 2024
@slowli @gabrieldemian
refactor(db): Combine storage log pruning into single query (matter-l…
d838f22 
…abs#2279)

## What ❔

Minor follow-up for matter-labs#2268
that combines both parts of log pruning into a single query.

## Why ❔

Easier to maintain and could be slightly more efficient since
intermediate data doesn't need to travel from Postgres to the node and
back.

## Checklist

- [x] PR title corresponds to the body of PR (we generate changelog
entries from PRs).
- [x] Documentation comments have been added / updated.
- [x] Code has been formatted via `zk fmt` and `zk lint`.
- [x] Spellcheck has been run via `zk spellcheck`.
github-merge-queue bot pushed a commit that referenced this pull request Jun 24, 2024
@zksync-era-bot
chore(main): release core 24.8.0 (#2135)
9080452 
🤖 I have created a release *beep* *boop*
---


##
[24.8.0](core-v24.7.0...core-v24.8.0)
(2024-06-24)


### ⚠ BREAKING CHANGES

* updated boojum and nightly rust compiler
([#2126](#2126))

### Features

* Add metrics for transaction execution result in state keeper
([#2021](#2021))
([dde0fc4](dde0fc4))
* **api:** Add new `l1_committed` block tag
([#2282](#2282))
([d5e8e9b](d5e8e9b))
* **api:** Rework zks_getProtocolVersion
([#2146](#2146))
([800b8f4](800b8f4))
* change `zkSync` occurences to `ZKsync`
([#2227](#2227))
([0b4104d](0b4104d))
* **contract-verifier:** Adjust contract verifier for zksolc 1.5.0
([#2255](#2255))
([63efb2e](63efb2e))
* **docs:** Add documentation for subset of wiring layer
implementations, used by Main node
([#2292](#2292))
([06c287b](06c287b))
* **docs:** Pruning and Snapshots recovery basic docs
([#2265](#2265))
([619a525](619a525))
* **en:** Allow recovery from specific snapshot
([#2137](#2137))
([ac61fed](ac61fed))
* **eth-sender:** fix for missing eth_txs_history entries
([#2236](#2236))
([f05b0ae](f05b0ae))
* Expose fair_pubdata_price for blocks and batches
([#2244](#2244))
([0d51cd6](0d51cd6))
* **merkle-tree:** Rework tree rollback
([#2207](#2207))
([c3b9c38](c3b9c38))
* **node-framework:** Add Main Node Client layer
([#2132](#2132))
([927d842](927d842))
* **node:** Move some stuff around
([#2151](#2151))
([bad5a6c](bad5a6c))
* **node:** Port (most of) Node to the Node Framework
([#2196](#2196))
([7842bc4](7842bc4))
* **object-store:** Allow caching object store objects locally
([#2153](#2153))
([6c6e65c](6c6e65c))
* **proof_data_handler:** add new endpoints to the TEE prover interface
API ([#1993](#1993))
([eca98cc](eca98cc))
* **prover:** Add file based config for fri prover gateway
([#2150](#2150))
([81ffc6a](81ffc6a))
* Remove initialize_components function
([#2284](#2284))
([0a38891](0a38891))
* **state-keeper:** Add metric for l2 block seal reason
([#2229](#2229))
([f967e6d](f967e6d))
* **state-keeper:** More state keeper metrics
([#2224](#2224))
([1e48cd9](1e48cd9))
* **sync-layer:** adapt MiniMerkleTree to manage priority queue
([#2068](#2068))
([3e72364](3e72364))
* **tee_verifier_input_producer:** use
`FactoryDepsDal::get_factory_deps()
([#2271](#2271))
([2c0a00a](2c0a00a))
* **toolbox:** add zk_toolbox ci
([#1985](#1985))
([4ab4922](4ab4922))
* updated boojum and nightly rust compiler
([#2126](#2126))
([9e39f13](9e39f13))
* upgraded encoding of transactions in consensus Payload.
([#2245](#2245))
([cb6a6c8](cb6a6c8))
* Use info log level for crates named zksync_* by default
([#2296](#2296))
([9303142](9303142))
* verification of L1Batch witness (BFT-471) - attempt 2
([#2232](#2232))
([dbcf3c6](dbcf3c6))
* verification of L1Batch witness (BFT-471)
([#2019](#2019))
([6cc5455](6cc5455))
* **vm-runner:** add basic metrics
([#2203](#2203))
([dd154f3](dd154f3))
* **vm-runner:** add protective reads persistence flag for state keeper
([#2307](#2307))
([36d2eb6](36d2eb6))
* **vm-runner:** shadow protective reads using VM runner
([#2017](#2017))
([1402dd0](1402dd0))


### Bug Fixes

* **api:** Fix getting pending block
([#2186](#2186))
([93315ba](93315ba))
* **api:** Fix transaction methods for pruned transactions
([#2168](#2168))
([00c4cca](00c4cca))
* **config:** Fix object store
([#2183](#2183))
([551cdc2](551cdc2))
* **config:** Split object stores
([#2187](#2187))
([9bcdabc](9bcdabc))
* **db:** Fix `insert_proof_generation_details()`
([#2291](#2291))
([c2412cf](c2412cf))
* **db:** Optimize `get_l2_blocks_to_execute_for_l1_batch`
([#2199](#2199))
([06ec5f3](06ec5f3))
* **en:** Fix reorg detection in presence of tree data fetcher
([#2197](#2197))
([20da566](20da566))
* **en:** Fix transient error detection in consistency checker
([#2140](#2140))
([38fdfe0](38fdfe0))
* **en:** Remove L1 client health check
([#2136](#2136))
([49198f6](49198f6))
* **eth-sender:** Don't resend already sent transactions in the same
block ([#2208](#2208))
([3538e9c](3538e9c))
* **eth-sender:** etter error handling in eth-sender
([#2163](#2163))
([0cad504](0cad504))
* **node_framework:** Run gas adjuster task only if necessary
([#2266](#2266))
([2dac846](2dac846))
* **object-store:** Consider more GCS errors transient
([#2246](#2246))
([2f6cd41](2f6cd41))
* **prover_cli:** Remove outdated fix for circuit id in node wg
([#2248](#2248))
([db8e71b](db8e71b))
* **prover:** Disallow state changes from successful
([#2233](#2233))
([2488a76](2488a76))
* **pruning:** Check pruning in metadata calculator
([#2286](#2286))
([7bd8f27](7bd8f27))
* Treat 502s and 503s as transient for GCS OS
([#2202](#2202))
([0a12c52](0a12c52))
* **vm-runner:** add config value for the first processed batch
([#2158](#2158))
([f666717](f666717))
* **vm-runner:** make `last_ready_batch` account for
`first_processed_batch`
([#2238](#2238))
([3889794](3889794))
* **vm:** fix insertion to `decommitted_code_hashes`
([#2275](#2275))
([15bb71e](15bb71e))
* **vm:** Update `decommitted_code_hashes` in `prepare_to_decommit`
([#2253](#2253))
([6c49a50](6c49a50))


### Performance Improvements

* **db:** Improve storage switching for state keeper cache
([#2234](#2234))
([7c8e24c](7c8e24c))
* **db:** Try yet another storage log pruning approach
([#2268](#2268))
([3ee34be](3ee34be))
* **en:** Parallelize persistence and chunk processing during tree
recovery
([#2050](#2050))
([b08a667](b08a667))
* **pruning:** Use more efficient query to delete past storage logs
([#2179](#2179))
([4c18755](4c18755))


### Reverts

* **pruning:** Revert pruning query
([#2220](#2220))
([8427cdd](8427cdd))
* verification of L1Batch witness (BFT-471)
([#2230](#2230))
([227e101](227e101))

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

---------

Co-authored-by: zksync-era-bot <zksync-era-bot@users.noreply.github.com>
irnb pushed a commit to vianetwork/via-server that referenced this pull request Jul 12, 2024
@zksync-era-bot @irnb
chore(main): release core 24.8.0 (matter-labs#2135)
20fd084 
🤖 I have created a release *beep* *boop*
---


##
[24.8.0](matter-labs/zksync-era@core-v24.7.0...core-v24.8.0)
(2024-06-24)


### ⚠ BREAKING CHANGES

* updated boojum and nightly rust compiler
([matter-labs#2126](matter-labs#2126))

### Features

* Add metrics for transaction execution result in state keeper
([matter-labs#2021](matter-labs#2021))
([dde0fc4](matter-labs@dde0fc4))
* **api:** Add new `l1_committed` block tag
([matter-labs#2282](matter-labs#2282))
([d5e8e9b](matter-labs@d5e8e9b))
* **api:** Rework zks_getProtocolVersion
([matter-labs#2146](matter-labs#2146))
([800b8f4](matter-labs@800b8f4))
* change `zkSync` occurences to `ZKsync`
([matter-labs#2227](matter-labs#2227))
([0b4104d](matter-labs@0b4104d))
* **contract-verifier:** Adjust contract verifier for zksolc 1.5.0
([matter-labs#2255](matter-labs#2255))
([63efb2e](matter-labs@63efb2e))
* **docs:** Add documentation for subset of wiring layer
implementations, used by Main node
([matter-labs#2292](matter-labs#2292))
([06c287b](matter-labs@06c287b))
* **docs:** Pruning and Snapshots recovery basic docs
([matter-labs#2265](matter-labs#2265))
([619a525](matter-labs@619a525))
* **en:** Allow recovery from specific snapshot
([matter-labs#2137](matter-labs#2137))
([ac61fed](matter-labs@ac61fed))
* **eth-sender:** fix for missing eth_txs_history entries
([matter-labs#2236](matter-labs#2236))
([f05b0ae](matter-labs@f05b0ae))
* Expose fair_pubdata_price for blocks and batches
([matter-labs#2244](matter-labs#2244))
([0d51cd6](matter-labs@0d51cd6))
* **merkle-tree:** Rework tree rollback
([matter-labs#2207](matter-labs#2207))
([c3b9c38](matter-labs@c3b9c38))
* **node-framework:** Add Main Node Client layer
([matter-labs#2132](matter-labs#2132))
([927d842](matter-labs@927d842))
* **node:** Move some stuff around
([matter-labs#2151](matter-labs#2151))
([bad5a6c](matter-labs@bad5a6c))
* **node:** Port (most of) Node to the Node Framework
([matter-labs#2196](matter-labs#2196))
([7842bc4](matter-labs@7842bc4))
* **object-store:** Allow caching object store objects locally
([matter-labs#2153](matter-labs#2153))
([6c6e65c](matter-labs@6c6e65c))
* **proof_data_handler:** add new endpoints to the TEE prover interface
API ([matter-labs#1993](matter-labs#1993))
([eca98cc](matter-labs@eca98cc))
* **prover:** Add file based config for fri prover gateway
([matter-labs#2150](matter-labs#2150))
([81ffc6a](matter-labs@81ffc6a))
* Remove initialize_components function
([matter-labs#2284](matter-labs#2284))
([0a38891](matter-labs@0a38891))
* **state-keeper:** Add metric for l2 block seal reason
([matter-labs#2229](matter-labs#2229))
([f967e6d](matter-labs@f967e6d))
* **state-keeper:** More state keeper metrics
([matter-labs#2224](matter-labs#2224))
([1e48cd9](matter-labs@1e48cd9))
* **sync-layer:** adapt MiniMerkleTree to manage priority queue
([matter-labs#2068](matter-labs#2068))
([3e72364](matter-labs@3e72364))
* **tee_verifier_input_producer:** use
`FactoryDepsDal::get_factory_deps()
([matter-labs#2271](matter-labs#2271))
([2c0a00a](matter-labs@2c0a00a))
* **toolbox:** add zk_toolbox ci
([matter-labs#1985](matter-labs#1985))
([4ab4922](matter-labs@4ab4922))
* updated boojum and nightly rust compiler
([matter-labs#2126](matter-labs#2126))
([9e39f13](matter-labs@9e39f13))
* upgraded encoding of transactions in consensus Payload.
([matter-labs#2245](matter-labs#2245))
([cb6a6c8](matter-labs@cb6a6c8))
* Use info log level for crates named zksync_* by default
([matter-labs#2296](matter-labs#2296))
([9303142](matter-labs@9303142))
* verification of L1Batch witness (BFT-471) - attempt 2
([matter-labs#2232](matter-labs#2232))
([dbcf3c6](matter-labs@dbcf3c6))
* verification of L1Batch witness (BFT-471)
([matter-labs#2019](matter-labs#2019))
([6cc5455](matter-labs@6cc5455))
* **vm-runner:** add basic metrics
([matter-labs#2203](matter-labs#2203))
([dd154f3](matter-labs@dd154f3))
* **vm-runner:** add protective reads persistence flag for state keeper
([matter-labs#2307](matter-labs#2307))
([36d2eb6](matter-labs@36d2eb6))
* **vm-runner:** shadow protective reads using VM runner
([matter-labs#2017](matter-labs#2017))
([1402dd0](matter-labs@1402dd0))


### Bug Fixes

* **api:** Fix getting pending block
([matter-labs#2186](matter-labs#2186))
([93315ba](matter-labs@93315ba))
* **api:** Fix transaction methods for pruned transactions
([matter-labs#2168](matter-labs#2168))
([00c4cca](matter-labs@00c4cca))
* **config:** Fix object store
([matter-labs#2183](matter-labs#2183))
([551cdc2](matter-labs@551cdc2))
* **config:** Split object stores
([matter-labs#2187](matter-labs#2187))
([9bcdabc](matter-labs@9bcdabc))
* **db:** Fix `insert_proof_generation_details()`
([matter-labs#2291](matter-labs#2291))
([c2412cf](matter-labs@c2412cf))
* **db:** Optimize `get_l2_blocks_to_execute_for_l1_batch`
([matter-labs#2199](matter-labs#2199))
([06ec5f3](matter-labs@06ec5f3))
* **en:** Fix reorg detection in presence of tree data fetcher
([matter-labs#2197](matter-labs#2197))
([20da566](matter-labs@20da566))
* **en:** Fix transient error detection in consistency checker
([matter-labs#2140](matter-labs#2140))
([38fdfe0](matter-labs@38fdfe0))
* **en:** Remove L1 client health check
([matter-labs#2136](matter-labs#2136))
([49198f6](matter-labs@49198f6))
* **eth-sender:** Don't resend already sent transactions in the same
block ([matter-labs#2208](matter-labs#2208))
([3538e9c](matter-labs@3538e9c))
* **eth-sender:** etter error handling in eth-sender
([matter-labs#2163](matter-labs#2163))
([0cad504](matter-labs@0cad504))
* **node_framework:** Run gas adjuster task only if necessary
([matter-labs#2266](matter-labs#2266))
([2dac846](matter-labs@2dac846))
* **object-store:** Consider more GCS errors transient
([matter-labs#2246](matter-labs#2246))
([2f6cd41](matter-labs@2f6cd41))
* **prover_cli:** Remove outdated fix for circuit id in node wg
([matter-labs#2248](matter-labs#2248))
([db8e71b](matter-labs@db8e71b))
* **prover:** Disallow state changes from successful
([matter-labs#2233](matter-labs#2233))
([2488a76](matter-labs@2488a76))
* **pruning:** Check pruning in metadata calculator
([matter-labs#2286](matter-labs#2286))
([7bd8f27](matter-labs@7bd8f27))
* Treat 502s and 503s as transient for GCS OS
([matter-labs#2202](matter-labs#2202))
([0a12c52](matter-labs@0a12c52))
* **vm-runner:** add config value for the first processed batch
([matter-labs#2158](matter-labs#2158))
([f666717](matter-labs@f666717))
* **vm-runner:** make `last_ready_batch` account for
`first_processed_batch`
([matter-labs#2238](matter-labs#2238))
([3889794](matter-labs@3889794))
* **vm:** fix insertion to `decommitted_code_hashes`
([matter-labs#2275](matter-labs#2275))
([15bb71e](matter-labs@15bb71e))
* **vm:** Update `decommitted_code_hashes` in `prepare_to_decommit`
([matter-labs#2253](matter-labs#2253))
([6c49a50](matter-labs@6c49a50))


### Performance Improvements

* **db:** Improve storage switching for state keeper cache
([matter-labs#2234](matter-labs#2234))
([7c8e24c](matter-labs@7c8e24c))
* **db:** Try yet another storage log pruning approach
([matter-labs#2268](matter-labs#2268))
([3ee34be](matter-labs@3ee34be))
* **en:** Parallelize persistence and chunk processing during tree
recovery
([matter-labs#2050](matter-labs#2050))
([b08a667](matter-labs@b08a667))
* **pruning:** Use more efficient query to delete past storage logs
([matter-labs#2179](matter-labs#2179))
([4c18755](matter-labs@4c18755))


### Reverts

* **pruning:** Revert pruning query
([matter-labs#2220](matter-labs#2220))
([8427cdd](matter-labs@8427cdd))
* verification of L1Batch witness (BFT-471)
([matter-labs#2230](matter-labs#2230))
([227e101](matter-labs@227e101))

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

---------

Co-authored-by: zksync-era-bot <zksync-era-bot@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@tomg10 tomg10 tomg10 approved these changes

@RomanBrodetski RomanBrodetski Awaiting requested review from RomanBrodetski

@perekopskiy perekopskiy Awaiting requested review from perekopskiy

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@slowli @tomg10