Create 16 byte _tsids in TsidBuilder #133706
Draft
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
Based on testing, this has marginal impact on storage but decreases the size of the _tsids from 21B-36B to 18B-21B.
elasticsearchmachine
added
external-contributor
3 tasks
|
Unfortunately, this seems to have a big negative impact of Full Rally comparison ...Track params ... |
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Outdated suggestions cannot be applied.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.
A variation of #133631 that creates even smaller _tsids, fixed at 16 bytes.
The impact on index size is marginal but has the potential to speed up querying.
One tradeoff is that it's lightly more prone to collisions, as this is using 112 bits instead of 128 bits for the "uniqueness portion" of the tsid.
According to the Birthday attack simple approximation formula, it would take about 100 trillion time series to have a one in a million chance for a collision:$\sqrt{2^{1+112-20}}$ .
While we need to account for the hash not being perfectly balanced, this should be safe.
However, we should test this also on non-OTel datasets.