Introduce bifrost benchpress #1545
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
This was referenced May 22, 2024
AhmedSoliman
force-pushed
the
pr1545
branch
2 times, most recently
from
fdb7bb8
to
1200e4c
Compare
AhmedSoliman
force-pushed
the
pr1545
branch
2 times, most recently
from
f34d8ee
to
e263662
Compare
tillrohrmann
approved these changes
May 23, 2024
There was a problem hiding this comment.
Great work @AhmedSoliman. The benchpress looks strong to me :-) +1 for merging.
This PR changes the default configuration of the local loglet to disable time-based batching. Most users's first impressions and general use will expect interactive low-latency experience and we want this to be the default case. A major performance win is the change in local loglet implementation. Tokio has 1ms timer resolution, so even if batching duration is set to zero, the runtime will still sleep an arbitrary amound between 0 and 1ms to resume the stream. The change switches to chunking ready items by default if time-based batching is zero. This results in 6X lower append path latency in my tests (even with fdatasync running on fast nvme)
We introduce bifrost benchpress, a binary that's used to run simulation perf tests focused on bifrost loglets performance. Initial code contains tests to reliably measure local loglet's critical-path single log append latency and write-to-read latency with a given configuration. The benchmark utility is built to give full control over how the benchmark runs, in contrast to criterion which is designed for micro benches.
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Outdated suggestions cannot be applied.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.
Introduce bifrost benchpress
We introduce bifrost benchpress, a binary that's used to run simulation perf tests focused on bifrost loglets performance. Initial code contains tests to reliably measure local loglet's critical-path single log append latency and write-to-read latency with a given configuration.
The benchmark utility is built to give full control over how the benchmark runs, in contrast to criterion which is designed for micro benches.
Stack created with Sapling. Best reviewed with ReviewStack.