Voting is open for a new IC release - 4e51609

Hello there!

we are happy to announce that voting is now open for a new IC release.
The NNS proposal is here: IC NNS Proposal 71669.

There are lots of exciting new features and bugfixes. Here is a more complete list of changes:

* Consensus: Add canister http metrics
* Consensus: Canister Http Delivery Fix
* Consensus: Do a size check in `validate_ingress`
* Consensus: Ensure ecdsa payload is empty when upgrading
* Consensus: Fix ecdsa response delivery to not depend on state
* Consensus: Fix small bug found in batch.rs
* Consensus: Remove evaludate_max_rps from workload generator
* Crypto: Add bls12_381 to MIRACL conversion functions
* Crypto: Add support for G2Prepared and multipairings in bls12_381 wrapper
* Crypto: Add support for comparison of BLS12-381 Scalars
* Crypto: Create corruption functions for Threshold ECDSA types
* Crypto: Remove obsolete potential error from multi-sig sign()-API
* Crypto: Remove unnecessary `mut` in CSP key generation API
* Crypto: Switch to using zkcrypto for hash to curve G2
* Crypto: Upgrade zkcrypto/bls12_381 to 0.7.0 to benefit from performance improvements
* Execution: Provisional apis should refund cycles correctly
* Execution: Reject queries to frozen canisters
* Execution: propagate subnet size to execution
* Message Routing: Incrementally compute manifest for WASM binaries
* Message Routing: Make checkpoint deletion more robust
* Message Routing: Persist wasm module hash in checkpoint
* Networking: Acquire the RwLock of the peer_map before on_connect_setup
* Networking: Add IntGaugeResource for keeping track of tokio tasks
* Networking: Add reasoning why we won't remove enforce_https for now
* Networking: Block replica ports that are used byto prevent users from accessing the replica
* Networking: Fix the broken gauges for read and write task tracking
* Networking: Improve btc test
* Networking: Improve logging and introduce tcp_accepts_total and tcp_connects_total metrics
* Networking: Make flow_write_task and flow_read_task be &self instead of associated functions
* Networking: Make some of the FlowState data members private
* Networking: Remove the logic for the "sender error" bit in the header
* Networking: Small refactoring of transport code
* Networking: Update the ConnectionState on_disconnect only after we send the event via the the event_handler
* Node: Revert DNS based NNS_URL until boundary-nodes fully support IPv6
* Node: Slow down node registration to give HSM time
* Node: Update build sha for snp image
* Node: Use DNS based NNS_URL instead of hard-coded node addresses
* Node: Use new baseimages in build
* Orchestrator: Better logs for replica
* Orchestrator: Fetch CUPs from more peers until one is successful ...
* Orchestrator: Simplify the polling delay between registry fetches
* Orchestrator: Update NNS URLs at each iteration of node registration
* Runtime: Add a config for slice instruction limit
* Runtime: Add metrics for module deserialization
* Runtime: Add wasm deserialization benchmarks
* Runtime: Enable module sharing
* Runtime: Fix DTS instruction bookkeeping in Wasm executor
* Runtime: Improve DTS out-of-instructions handler
* Runtime: Make ExecuteMessageResult an enum for DTS
* Runtime: Move round instruction limit to `RoundLimits`
* Runtime: Propagate slice executed instructions from Wasm
* Runtime: Refactor instruction limits in `ExecutionParameters`
* Runtime: Remove cycles u128 flag
* Runtime: Remove support for changing compute allocation during install
* Runtime: Remove unneeded return in stable-structures
* Runtime: Resolves TODOs in `SandboxedExecutionController`
* Runtime: Simplify DTS for install code
* Runtime: Track compiled canisters in ReplicatedState
* Runtime: Use the `CompilationCostHandling` when adjusting round limits
* Various tech-debt management: code refactoring, docs, bug fixes, test updates

And a complete list of changes can of course be found on GitHub.
Please reply to this message if you have any questions or comments.

4 Likes

Hello everyone!

As last week, there is a proposal out to elect a new IC replica release which enablea the separation of the Crypto Service Provider (CSP) into a separate process. To make sure this change does not affect the availability of the IC or introduce any regressions, the plan would be to roll this out slowly to the subnets, and therefore it could take a few release cycles. Since last week’s release was successful, this week it will be rolled out to more subnets.

The proposal can found here .
Please do reach out if you have any questions about it.

All the best,
Andrea