ring buffer: store logs as proto instead of fb #297
+867
−883
Conversation
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
mattklein123
force-pushed
the
proto
branch
2 times, most recently
from
cefd6fe to
11be406
Compare
Closed
For logs with many fields, the fact that fb is padded and doesn't use varint encoding leads to substantial overhead. This removes that overhead. As part of this change I went a bit over the top and also implemented a custom encoder for the Log proto which allows for zero copy into the ring buffer. This is better than the previous fb implementation which encoded and then did a memcpy into the buffer. Signed-off-by: Matt Klein <[email protected]>
snowp
reviewed
Oct 31, 2025
| // Helpers for doing raw casts where we are sure the value fits and don't want to pay for | ||
| // checks and avoid clippy lints. | ||
| pub trait LossyIntToU32 { | ||
| fn to_u32(self) -> u32; |
There was a problem hiding this comment.
should this be called to_u32_lossy just to make it obvious?
| ) -> u64 { | ||
| let mut my_size = 0; | ||
|
|
||
| my_size += ::protobuf::rt::uint64_size(1, occurred_at.unix_timestamp_nanos().to_u64() / 1000); |
There was a problem hiding this comment.
where does nanos / 1000 come from?
There was a problem hiding this comment.
It's microseconds since epoch. I will add more comments everywhere.
| ) | ||
| } | ||
|
|
||
| pub fn serialize_proto_to_stream_inner( |
There was a problem hiding this comment.
Obviously this means that we need to maintain this as more fields are added - any way to have a test fail in that case?
| os.write_uint32(2, log_level)?; | ||
| } | ||
|
|
||
| Self::serialize_proto_data(3, message, os)?; |
There was a problem hiding this comment.
I assumed you checked but there's no constants available in the generated proto with the field IDs?
There was a problem hiding this comment.
I don't think so but will double check.
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.
For logs with many fields, the fact that fb is padded and doesn't use varint encoding leads to substantial overhead. This removes that overhead. As part of this change I went a bit over the top and also implemented a custom encoder for the Log proto which allows for zero copy into the ring buffer. This is better than the previous fb implementation which encoded and then did a memcpy into the buffer.