Performance improvement after output refactoring#89
Open
dpirotte wants to merge 1 commit intoeulerto:masterfrom
Open
Performance improvement after output refactoring#89dpirotte wants to merge 1 commit intoeulerto:masterfrom
dpirotte wants to merge 1 commit intoeulerto:masterfrom
Conversation
Removing pretty-print branching replaced several appendStringInfoString uses with appendStringInfo, which dramatically cleaned up the output code at the cost of decoding about 22% slower. We can recover most of this performance and keep the tidiness by adding a helper function to run appendStringInfoString for a variable number of strings. We lose some niceties of the sprintf-style template, but the performance seems worth it for high-throughput databases. "22% slower" was measured by populating a pgbench database with scale factor 5, running 50K transactions, DELETE-ing all the data, then timing how long it takes to receive all the changes with `pg_recvlogical`. On a c5.4xlarge EC2 instance backed by a gp2 EBS volume, this takes about 9.2 seconds on this branch compared to about 11.2s on master as of aa4077a.
Contributor
Author
|
I ran some additional benchmarks overnight with the same script to verify the performance delta, and it remains consistent.
|
|
@eulerto: this looks interesting. Thoughts? |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
In an attempt to bring our fork closer to master, we discovered that d4c0e81 introduced a meaningful performance change. d4c0e81 replaced several appendStringInfoString uses with appendStringInfo, which dramatically cleaned up the output code at the cost of a performance hit.
This proposal reduces decoding time by ~20% as measured by this benchmarking script. It swaps
appendStringInfofor a newappendStringInfoStringshelper, which loops through a variable argument list and calls the more performantappendStringInfoStringunder the hood, skipping if the argument starts with a null byte, i.e. whitespace args whenpretty-printisn't set. (I think this is safe in these use cases, and it is responsible for a full quarter of the performance return.)I'm very open to suggestions or other approaches that keep the (very nice) output refactoring and reclaim most of the performance.