perf: evolve promql execution engine #5691

waynexia · 2025-03-12T00:07:37Z

Signed-off-by: Ruihang Xia [email protected]I hereby agree to the terms of the GreptimeDB CLA.

Refer to a related PR or issue link (optional)

What's changed and what's your intention?

100x performance improvement and 40x memory reduction by eliminating all potential sorts, and parallelizing as much as possible.

PR Checklist

Please convert it to a draft if some of the following conditions are not met.

I have written the necessary rustdoc comments.
I have added the necessary unit tests and integration tests.
This PR requires documentation updates.
API changes are backward compatible.
Schema or data changes are backward compatible.

Signed-off-by: Ruihang Xia <[email protected]>

This reverts commit db58884.

Signed-off-by: Ruihang Xia <[email protected]>

coderabbitai · 2025-03-12T00:07:43Z

Important

Review skipped

Auto reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Generate unit testing code for this file.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai generate unit testing code for this file.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and generate unit testing code.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

Signed-off-by: Ruihang Xia <[email protected]>

killme2008

Good job!

killme2008 · 2025-03-12T03:33:33Z

src/query/src/optimizer/pass_distribution.rs

+            }
+
+            if let Some(merge_scan) = plan.as_any().downcast_ref::<MergeScanExec>()
+                && let Some(distribution) = distribution_requirement.as_ref()


Looks like we can take the distribution_requirement and avoid the below clone

killme2008 · 2025-03-12T03:37:09Z

src/query/src/dist_plan/merge_scan.rs

-            ExecutionMode::Bounded,
-        );
+
+        let eq_properties = if let LogicalPlan::Sort(sort) = &plan


Could you add some commments to explain what's happening here.

evenyag · 2025-03-13T12:42:58Z

src/promql/src/extension_plan/series_divide.rs

@@ -375,11 +392,34 @@ impl SeriesDivideStream {
                    let current_value = current_string_array.value(0);
                    let last_value = last_string_array.value(last_row);
                    if current_value != last_value {
-                        return Ok(Some((resumed_batch_index, 0)));
+                        return Ok(Some((resumed_batch_index - 1, last_batch.num_rows() - 1)));


Do we have test cases for the changes in this file?

evenyag · 2025-03-13T12:57:20Z

src/query/src/promql/planner.rs

@@ -376,7 +376,7 @@ impl PromPlanner {

        let group_sort_expr = new_group_exprs
            .into_iter()
-            .map(|expr| expr.sort(true, false));
+            .map(|expr| expr.sort(false, false));


Why do we change this to DESC? It changes topk/bottomk output order.

evenyag · 2025-03-13T13:24:43Z

tests/cases/standalone/common/promql/scalar.result

-| 1970-01-01T00:00:05 | 7.0                               |
-| 1970-01-01T00:00:10 | 11.0                              |
-| 1970-01-01T00:00:15 | 15.0                              |
+| 1970-01-01T00:00:00 | NaN                               |


Is the result correct?

evenyag · 2025-03-13T13:30:55Z

tests/cases/standalone/common/types/float/nan_join.result

++
++


Do we need to fix this?

evenyag · 2025-03-13T13:31:40Z

tests/cases/standalone/common/promql/topk_bottomk.result

+| 5   | host2 | idc1 | 1970-01-01T00:00:10 |
+| 4   | host2 | idc1 | 1970-01-01T00:00:05 |
+| 3   | host3 | idc2 | 1970-01-01T00:00:00 |


Should we sort results by time asc?

evenyag · 2025-03-13T13:33:00Z

Cargo.toml

@@ -129,7 +129,7 @@ etcd-client = "0.14"
 fst = "0.4.7"
 futures = "0.3"
 futures-util = "0.3"
-greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "c5419bbd20cb42e568ec325a4d71a3c94cc327e1" }
+greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "f01d022451be031bcfb00c414939ecbf059e80ea" }


Wait for GreptimeTeam/greptime-proto#223

Signed-off-by: Ruihang Xia <[email protected]>

waynexia added 16 commits March 5, 2025 08:30

use the same sort option across every prom plan

b4ae9b0

Signed-off-by: Ruihang Xia <[email protected]>

tweak plans

c425be2

Signed-off-by: Ruihang Xia <[email protected]>

wip

db58884

Signed-off-by: Ruihang Xia <[email protected]>

Merge branch 'main' into perf-prom-plan

270303e

Signed-off-by: Ruihang Xia <[email protected]>

fix merge compile

708e6b1

Signed-off-by: Ruihang Xia <[email protected]>

Revert "wip"

ef85167

This reverts commit db58884.

tweak merge scan

81d122b

Signed-off-by: Ruihang Xia <[email protected]>

handle error

99ec5cf

Signed-off-by: Ruihang Xia <[email protected]>

pass distribution rule

7119497

Signed-off-by: Ruihang Xia <[email protected]>

reverse sort order

99241a0

Signed-off-by: Ruihang Xia <[email protected]>

refine plans

d88b7ed

Signed-off-by: Ruihang Xia <[email protected]>

Merge branch 'main' into perf-prom-plan

23b9e16

Signed-off-by: Ruihang Xia <[email protected]>

more optimizations for plans

4bc08bc

Signed-off-by: Ruihang Xia <[email protected]>

Merge branch 'main' into perf-prom-plan

ef2cd42

Signed-off-by: Ruihang Xia <[email protected]>

check logical table

9bd1b84

Signed-off-by: Ruihang Xia <[email protected]>

fix tests

dad0f6c

Signed-off-by: Ruihang Xia <[email protected]>

waynexia requested review from evenyag and a team as code owners March 12, 2025 00:07

github-actions bot added the docs-not-required This change does not impact docs. label Mar 12, 2025

waynexia force-pushed the perf-prom-plan branch from 867dace to 93f589d Compare March 12, 2025 00:57

wierd tests

38ffd3a

Signed-off-by: Ruihang Xia <[email protected]>

waynexia force-pushed the perf-prom-plan branch from 93f589d to 38ffd3a Compare March 12, 2025 01:18

killme2008 reviewed Mar 12, 2025

View reviewed changes

evenyag reviewed Mar 13, 2025

View reviewed changes

Merge branch 'main' into perf-prom-plan

5550d4d

Signed-off-by: Ruihang Xia <[email protected]>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

perf: evolve promql execution engine #5691

perf: evolve promql execution engine #5691

waynexia commented Mar 12, 2025

coderabbitai bot commented Mar 12, 2025

Review skipped

Chat

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

killme2008 left a comment

killme2008 Mar 12, 2025

killme2008 Mar 12, 2025

evenyag Mar 13, 2025

evenyag Mar 13, 2025

evenyag Mar 13, 2025

evenyag Mar 13, 2025

evenyag Mar 13, 2025

evenyag Mar 13, 2025

perf: evolve promql execution engine #5691

Are you sure you want to change the base?

perf: evolve promql execution engine #5691

Conversation

waynexia commented Mar 12, 2025

Refer to a related PR or issue link (optional)

What's changed and what's your intention?

PR Checklist

coderabbitai bot commented Mar 12, 2025

Review skipped

Chat

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

killme2008 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment