This repository was archived by the owner on Mar 24, 2026. It is now read-only.
Only populate secondary endpoints if the config is set#7
Open
KenGordon wants to merge 2 commits into
Open
Conversation
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
DomAyre
force-pushed
the
fix-secondary-endpoints
branch
6 times, most recently
from
January 12, 2024 11:00
bdae18f to
6a2d1a3
Compare
Co-Authored-By: Dominic Ayre <dominicayre@microsoft.com>
DomAyre
force-pushed
the
fix-secondary-endpoints
branch
from
January 12, 2024 11:01
6a2d1a3 to
bfc96ba
Compare
This was referenced Jan 23, 2024
Open
jasarora-google
pushed a commit
that referenced
this pull request
Oct 12, 2024
### ⚠ BREAKING CHANGES * All sellers and buyers must update the js_num_workers parameter to udf_num_workers in terraform * Sellers must wait for all the integrated buyers to update to 4.1.0 before updating to accomodate serialization changes in the GetBidsRequest. ### Features * [BYOB] Add buyer code fetcher for executables * [BYOB] Add dispatch client for generateBid * [BYOB] Add GCP build setup for Bidding service * [BYOB] Add gRPC reactor for generateBid binary * [BYOB] Add info about runtime flags to demo/README.md * [BYOB] Add interface for clients that load code * [BYOB] Add to Bidding service critical path * Add a generic model validator interface for Tensorflow and PyTorch * Add PrivateAggregation protos to generateBid spec * add project setup scripts * add proto support for updateIfOlderThanMs * BFE parses perInterestGroupData, returns updateIfOlderThanMs data to SFE * Compress SFE <> BFE comms * Create new inference dashboard for GCP * Enable chaffing on SFE prod mode builds * Enable connecting to TKV via an External LB * Enable dynamic partition for inference metrics * Enable model validation for PyTorch * Enable TCMalloc for B&A inference * Export all available models as metric * Freeze a TensorFlow model ([bcc798c]( )), closes [#7]( ) * Implement a PyTorch model validator * Implement a TensorFlow model validator * Rename JS_NUM_WORKERS to UDF_NUM_WORKERS * SFE sends updateIfOlderThanMs data to client * update start_bidding with minimal defaults and local file flag ### Bug Fixes * BYOB generateBid should only parse first bid from binary response * Ensure gRPC calls don't hang on error * Fix js_execution.error.count metric * Fix memory leak in BFE client * Fix readonly declarations in shell script * Fix version bump in release script * Load request files in the init function * Reduce absl::GetFlag() call in GetBidsReactor * Renames prev_wins_ms to prev_wins * Set inference sidecar GRPC keepalive timeout * update tls cert commands for GCP setup * Use std::hash instead of absl::hash Bug: N/A Change-Id: Ia7d37c5a7e5be0e4c91c26c8fce4cc15683946a3 GitOrigin-RevId: 671dc23f8e04b711f7efe95907b66002e5a3b019
jasarora-google
added a commit
that referenced
this pull request
Oct 14, 2024
### ⚠ BREAKING CHANGES * All sellers and buyers must update the js_num_workers parameter to udf_num_workers in terraform * Sellers must wait for all the integrated buyers to update to 4.1.0 before updating to accomodate serialization changes in the GetBidsRequest. ### Features * [BYOB] Add buyer code fetcher for executables * [BYOB] Add dispatch client for generateBid * [BYOB] Add GCP build setup for Bidding service * [BYOB] Add gRPC reactor for generateBid binary * [BYOB] Add info about runtime flags to demo/README.md * [BYOB] Add interface for clients that load code * [BYOB] Add to Bidding service critical path * Add a generic model validator interface for Tensorflow and PyTorch * Add PrivateAggregation protos to generateBid spec * add project setup scripts * add proto support for updateIfOlderThanMs * BFE parses perInterestGroupData, returns updateIfOlderThanMs data to SFE * Compress SFE <> BFE comms * Create new inference dashboard for GCP * Enable chaffing on SFE prod mode builds * Enable connecting to TKV via an External LB * Enable dynamic partition for inference metrics * Enable model validation for PyTorch * Enable TCMalloc for B&A inference * Export all available models as metric * Freeze a TensorFlow model ([bcc798c]( )), closes [#7]( ) * Implement a PyTorch model validator * Implement a TensorFlow model validator * Rename JS_NUM_WORKERS to UDF_NUM_WORKERS * SFE sends updateIfOlderThanMs data to client * update start_bidding with minimal defaults and local file flag ### Bug Fixes * BYOB generateBid should only parse first bid from binary response * Ensure gRPC calls don't hang on error * Fix js_execution.error.count metric * Fix memory leak in BFE client * Fix readonly declarations in shell script * Fix version bump in release script * Load request files in the init function * Reduce absl::GetFlag() call in GetBidsReactor * Renames prev_wins_ms to prev_wins * Set inference sidecar GRPC keepalive timeout * update tls cert commands for GCP setup * Use std::hash instead of absl::hash Bug: N/A GitOrigin-RevId: 671dc23f8e04b711f7efe95907b66002e5a3b019 Change-Id: I16a2f5227cfdc86469fffcbd494c68b2b029a5ef
mihnjong-l
requested review from
jasarora-google
and removed request for
jasarora-google
December 11, 2024 18:37
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 subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
First change to support Azure port. It came about as a result of the Azure KMS implementation being different - there is only ever one KMS and so there is never a SECONDARY_COORDINATOR_PRIVATE_KEY_ENDPOINT.
This change does two main things:
Modifies
TrustedServersConfigClient::HasParameter()to return false when parameter is present in config_entries_map but has a default empty value.SECONDARY_COORDINATOR_PRIVATE_KEY_ENDPOINTisn't specified,TrustedServersConfigClient::HasParameter(SECONDARY_COORDINATOR_PRIVATE_KEY_ENDPOINT)incorrectly returns trueWhen initialising PrivateKeyFetcherFactory, reorganises the logic so that the secondary endpoint is only included when
SECONDARY_COORDINATOR_PRIVATE_KEY_ENDPOINTis setThese two changes mean that when combining primary and secondary endpoints into a single vector in the data plane, we don't get a secondary endpoint with a blank URL