Make FindMissingBlobs batch size configurable#1330
Open
bvinc wants to merge 1 commit into
Open
Conversation
The OSS RE client batches `FindMissingBlobs` requests at a hard-coded limit of 100 digests per RPC. With many digests this serializes into a long chain of small round-trips: 10k digests at ~300ms RTT becomes a 30-second wall-clock pause on the FMB stage. Expose the limit as `[buck2_re_client] find_missing_blobs_batch_size` (default 100, preserving existing behavior). Operators with deduped get_digests_ttl calls and a server that tolerates larger requests can raise this to collapse the FMB stage into a single round-trip. Larger values are only safe when paired with `[buck2] deduplicate_get_digests_ttl_calls = true`. Without dedup, N concurrent actions each issue an N-digest request simultaneously, producing a thundering herd against the RE backend; the existing 100 default was implicitly throttling that fanout via sequential batching.
meta-cla
Bot
added
the
CLA Signed
Contributor
|
This pull request has been imported. If you are a Meta employee, you can view this in D106717561. (Because this pull request was imported automatically, there will not be any future comments.) |
sluongng
pushed a commit
to sluongng/buck2
that referenced
this pull request
Jun 1, 2026
The OSS RE client used a fixed 100-digest limit for each FindMissingBlobs request. High-latency backends can spend a long time in serial round trips when an action has many digests to check. Expose the limit as buck2_re_client.find_missing_blobs_batch_size while preserving the 100-digest default. Clamp zero to one so an invalid value does not create a degenerate batching path. Original-Commit: 72aaf61 Source-PR: facebook#1330
sluongng
pushed a commit
to sluongng/buck2
that referenced
this pull request
Jun 3, 2026
The OSS RE client used a fixed 100-digest limit for each FindMissingBlobs request. High-latency backends can spend a long time in serial round trips when an action has many digests to check. Expose the limit as buck2_re_client.find_missing_blobs_batch_size while preserving the 100-digest default. Clamp zero to one so an invalid value does not create a degenerate batching path. Original-Commit: 72aaf61 Source-PR: facebook#1330
sluongng
pushed a commit
to sluongng/buck2
that referenced
this pull request
Jun 3, 2026
The OSS RE client used a fixed 100-digest limit for each FindMissingBlobs request. High-latency backends can spend a long time in serial round trips when an action has many digests to check. Expose the limit as buck2_re_client.find_missing_blobs_batch_size while preserving the 100-digest default. Clamp zero to one so an invalid value does not create a degenerate batching path. Original-Commit: 72aaf61 Source-PR: facebook#1330
sluongng
pushed a commit
to sluongng/buck2
that referenced
this pull request
Jun 11, 2026
The OSS RE client used a fixed 100-digest limit for each FindMissingBlobs request. High-latency backends can spend a long time in serial round trips when an action has many digests to check. Expose the limit as buck2_re_client.find_missing_blobs_batch_size while preserving the 100-digest default. Clamp zero to one so an invalid value does not create a degenerate batching path. Original-Commit: 72aaf61 Source-PR: facebook#1330
sluongng
pushed a commit
to sluongng/buck2
that referenced
this pull request
Jun 11, 2026
The OSS RE client used a fixed 100-digest limit for each FindMissingBlobs request. High-latency backends can spend a long time in serial round trips when an action has many digests to check. Expose the limit as buck2_re_client.find_missing_blobs_batch_size while preserving the 100-digest default. Clamp zero to one so an invalid value does not create a degenerate batching path. Original-Commit: 72aaf61 Source-PR: facebook#1330
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
1 participant
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.
The OSS RE client batches
FindMissingBlobsrequests at a hard-coded limit of 100 digests per RPC. With many digests this serializes into a long chain of small round-trips: 10k digests at ~300ms RTT becomes a 30-second wall-clock pause on the FMB stage.Expose the limit as
[buck2_re_client] find_missing_blobs_batch_size(default 100, preserving existing behavior). Operators with deduped get_digests_ttl calls and a server that tolerates larger requests can raise this to collapse the FMB stage into a single round-trip.Larger values are only safe when paired with
[buck2] deduplicate_get_digests_ttl_calls = true. Without dedup, N concurrent actions each issue an N-digest request simultaneously, producing a thundering herd against the RE backend; the existing 100 default was implicitly throttling that fanout via sequential batching.