Merge changes

Author: MauroToscano

.github/workflows/foundry-gas-diff.yml

@@ -35,7 +35,7 @@ jobs:
           FOUNDRY_FUZZ_SEED: 0x${{ github.event.pull_request.base.sha || github.sha }}
 
       - name: Compare gas reports
-        uses: Rubilmax/foundry-gas-diff@v3.16
+        uses: Rubilmax/foundry-gas-diff@v3.21
         with:
           summaryQuantile: 0.9 # only display the 10% most significant gas diffs in the summary (defaults to 20%)
           sortCriteria: avg,max # sort diff rows by criteria

.github/workflows/send-proofs-docker.yml

@@ -10,7 +10,7 @@ on:
       - '**.md'
 
 concurrency:
-  group: pull_request-${{ github.event.pull_request.number }}
+  group: ${{ github.event_name == 'merge_group' && format('merge_group-{0}', github.event.merge_group.head_sha) || format('pull_request-{0}', github.event.pull_request.number) }}
   cancel-in-progress: true
 
 jobs:

Makefile

@@ -7,7 +7,9 @@ CONFIG_FILE?=config-files/config.yaml
 export OPERATOR_ADDRESS ?= $(shell yq -r '.operator.address' $(CONFIG_FILE))
 AGG_CONFIG_FILE?=config-files/config-aggregator.yaml
 
-OPERATOR_VERSION=v0.13.0
+OPERATOR_VERSION=v0.14.0
+EIGEN_SDK_GO_VERSION_TESTNET=v0.2.0-beta.1
+EIGEN_SDK_GO_VERSION_MAINNET=v0.1.13
 
 ifeq ($(OS),Linux)
 	BUILD_ALL_FFI = $(MAKE) build_all_ffi_linux
@@ -30,6 +32,16 @@ ifeq ($(OS),Darwin)
 	BUILD_OPERATOR = $(MAKE) build_operator_macos
 endif
 
+ifeq ($(ENVIRONMENT), devnet)
+	GET_SDK_VERSION = $(MAKE) operator_set_eigen_sdk_go_version_devnet
+else ifeq ($(ENVIRONMENT), testnet)
+	GET_SDK_VERSION = $(MAKE) operator_set_eigen_sdk_go_version_testnet
+else ifeq ($(ENVIRONMENT), mainnet)
+	GET_SDK_VERSION = $(MAKE) operator_set_eigen_sdk_go_version_mainnet
+else
+	GET_SDK_VERSION = $(MAKE) operator_set_eigen_sdk_go_version_error
+endif
+
 
 FFI_FOR_RELEASE ?= true
 
@@ -140,7 +152,13 @@ anvil_start_with_block_time_with_more_prefunded_accounts:
 
 _AGGREGATOR_:
 
+build_aggregator:
+	$(GET_SDK_VERSION)
+	@echo "Building aggregator"
+	@go build -o ./build/aligned-aggregator ./aggregator/cmd/main.go
+
 aggregator_start:
+	$(GET_SDK_VERSION)
 	@echo "Starting Aggregator..."
 	@go run aggregator/cmd/main.go --config $(AGG_CONFIG_FILE) \
 	2>&1 | zap-pretty
@@ -156,15 +174,31 @@ test_go_retries:
 __OPERATOR__:
 
 operator_start:
+	$(GET_SDK_VERSION)
 	@echo "Starting Operator..."
 	go run operator/cmd/main.go start --config $(CONFIG_FILE) \
 	2>&1 | zap-pretty
 
+operator_set_eigen_sdk_go_version_testnet:
+	@echo "Setting Eigen SDK version to: $(EIGEN_SDK_GO_VERSION_TESTNET)"
+	go get github.com/Layr-Labs/eigensdk-go@$(EIGEN_SDK_GO_VERSION_TESTNET)
+
+operator_set_eigen_sdk_go_version_devnet: operator_set_eigen_sdk_go_version_mainnet
+
+operator_set_eigen_sdk_go_version_mainnet:
+	@echo "Setting Eigen SDK version to: $(EIGEN_SDK_GO_VERSION_MAINNET)"
+	go get github.com/Layr-Labs/eigensdk-go@$(EIGEN_SDK_GO_VERSION_MAINNET)
+
+operator_set_eigen_sdk_go_version_error:
+	@echo "Error setting Eigen SDK version, missing ENVIRONMENT. Possible values for ENVIRONMENT=<devnet|testnet|mainnet>"
+	exit 1
+
 operator_full_registration: operator_get_eth operator_register_with_eigen_layer operator_mint_mock_tokens operator_deposit_into_mock_strategy operator_whitelist_devnet operator_register_with_aligned_layer
 
-operator_register_and_start: operator_full_registration operator_start
+operator_register_and_start: $(GET_SDK_VERSION) operator_full_registration operator_start
 
 build_operator: deps
+	$(GET_SDK_VERSION)
 	$(BUILD_OPERATOR)
 
 build_operator_macos:
@@ -178,6 +212,7 @@ build_operator_linux:
 	@echo "Operator built into /operator/build/aligned-operator"
 
 update_operator:
+	$(GET_SDK_VERSION)
 	@echo "Updating Operator..."
 	@./scripts/fetch_latest_release.sh
 	@make build_operator

aggregator/pkg/server.go

@@ -3,6 +3,7 @@ package pkg
 import (
 	"context"
 	"encoding/hex"
+	"errors"
 	"fmt"
 	"net/http"
 	"net/rpc"
@@ -48,6 +49,17 @@ func (agg *Aggregator) ProcessOperatorSignedTaskResponseV2(signedTaskResponse *t
 		"SenderAddress", "0x"+hex.EncodeToString(signedTaskResponse.SenderAddress[:]),
 		"BatchIdentifierHash", "0x"+hex.EncodeToString(signedTaskResponse.BatchIdentifierHash[:]),
 		"operatorId", hex.EncodeToString(signedTaskResponse.OperatorId[:]))
+
+	if signedTaskResponse.BlsSignature.G1Point == nil {
+		agg.logger.Warn("invalid operator response with nil signature",
+			"BatchMerkleRoot", "0x"+hex.EncodeToString(signedTaskResponse.BatchMerkleRoot[:]),
+			"SenderAddress", "0x"+hex.EncodeToString(signedTaskResponse.SenderAddress[:]),
+			"BatchIdentifierHash", "0x"+hex.EncodeToString(signedTaskResponse.BatchIdentifierHash[:]),
+			"operatorId", hex.EncodeToString(signedTaskResponse.OperatorId[:]))
+		*reply = 1
+		return errors.New("invalid response: nil signature")
+	}
+
 	taskIndex := uint32(0)
 
 	// The Aggregator may receive the Task Identifier after the operators.
@@ -69,7 +81,7 @@ func (agg *Aggregator) ProcessOperatorSignedTaskResponseV2(signedTaskResponse *t
 	defer cancel() // Ensure the cancel function is called to release resources
 
 	// Create a channel to signal when the task is done
-	done := make(chan struct{})
+	done := make(chan uint8)
 
 	agg.logger.Info("Starting bls signature process")
 	go func() {
@@ -80,9 +92,11 @@ func (agg *Aggregator) ProcessOperatorSignedTaskResponseV2(signedTaskResponse *t
 
 		if err != nil {
 			agg.logger.Warnf("BLS aggregation service error: %s", err)
+			done<- 1
 			// todo shouldn't we here close the channel with a reply = 1?
 		} else {
 			agg.logger.Info("BLS process succeeded")
+			done<- 0
 		}
 
 		close(done)
@@ -94,10 +108,10 @@ func (agg *Aggregator) ProcessOperatorSignedTaskResponseV2(signedTaskResponse *t
 	case <-ctx.Done():
 		// The context's deadline was exceeded or it was canceled
 		agg.logger.Info("Bls process timed out, operator signature will be lost. Batch may not reach quorum")
-	case <-done:
+	case res := <-done:
 		// The task completed successfully
-		agg.logger.Info("Bls context finished correctly")
-		*reply = 0
+		agg.logger.Info("Bls context finished on time")
+		*reply = res
 	}
 
 	return nil

alerts/balance_alerts.sh

@@ -38,7 +38,7 @@ do
     fi
     balance_alert=true
   else
-    message="🟩 INFO: Wallet $WALLET_ADDRESS balance ($balance_eth ETH) is above $BALANCE_THRESHOLD ETH"
+    message="🟩 INFO: $WALLET_NAME ($NETWORK) Wallet $WALLET_ADDRESS balance ($balance_eth ETH) is above $BALANCE_THRESHOLD ETH"
     printf "$message\n"
     if [ "$balance_alert" = true ]; then
       send_slack_message "$message"

alerts/contract_alerts.sh

@@ -41,7 +41,8 @@ do
   if [ -z "$new_batch_logs" ]; then
     printf "No new batches logs found\n"
     if [ "$no_new_batches_alert" = false ]; then
-      message="🚨 ALERT: No new batches in Service Manager since block $from_block"
+      message="🚨 $NETWORK ALERT: No new batches in Service Manager since block $from_block"
+      printf "$message\n"
       send_slack_message "$message"
       send_telegram_message "$message"
       send_pagerduty_alert "$message"
@@ -50,7 +51,8 @@ do
   else
     printf "New batches logs found\n"
     if [ "$no_new_batches_alert" = true ]; then
-      message="🟩 INFO: Batches creation resumed in Service Manager since block $from_block"
+      message="🟩 $NETWORK INFO: Batches creation resumed in Service Manager since block $from_block"
+      printf "$message\n"
       send_slack_message "$message"
       send_telegram_message "$message"
     fi
@@ -61,7 +63,8 @@ do
   if [ -z "$verified_batch_logs" ]; then
     printf "No verified batches logs found\n"
     if [ "$no_verified_batches_alert" = false ]; then
-      message="🚨 ALERT: No verified batches in Service Manager since block $from_block"
+      message="🚨 $NETWORK ALERT: No verified batches in Service Manager since block $from_block"
+      printf "$message\n"
       send_slack_message "$message"
       send_telegram_message "$message"
       send_pagerduty_alert "$message"
@@ -70,7 +73,8 @@ do
   else
     printf "Verified batches logs found\n"
     if [ "$no_verified_batches_alert" = true ]; then
-      message="🟩 INFO: Batches verification resumed in Service Manager since block $from_block"
+      message="🟩 $NETWORK INFO: Batches verification resumed in Service Manager since block $from_block"
+      printf "$message\n"
       send_slack_message "$message"
       send_telegram_message "$message"
     fi

alerts/sender_with_alert.sh

@@ -92,8 +92,15 @@ do
 
   ## Generate Proof
   nonce=$(aligned get-user-nonce --network $NETWORK --user_addr $SENDER_ADDRESS 2>&1 | awk '{print $9}')
+  echo $nonce
+  if ! [[ "$nonce" =~ ^[0-9]+$ ]]; then
+    echo "Failed getting user nonce, retrying in 10 seconds"
+    sleep 10
+    continue
+  fi
+
   x=$((nonce + 1)) # So we don't have any issues with nonce = 0
-  echo "Generating proof $x != 0"
+  echo "Generating proof $x != 0, nonce: $nonce"
   go run ./scripts/test_files/gnark_groth16_bn254_infinite_script/cmd/main.go $x
 
   ## Send Proof
@@ -112,7 +119,26 @@ do
     2>&1)
 
   echo "$submit"
-  
+
+  submit_errors=$(echo "$submit" | grep -oE 'ERROR[^]]*]([^[]*)' | sed 's/^[^]]*]//;s/[[:space:]]*$//')
+
+  # Loop through each error found and print with the custom message
+  is_error=0
+  while IFS= read -r error; do
+      if [[ -n "$error" ]]; then
+          slack_error_message="Error submitting proof to $NETWORK: $error"
+          send_slack_message "$slack_error_message"
+          is_error=1
+      fi
+  done <<< "$submit_errors"
+
+  if [ $is_error -eq 1 ]; then
+    echo "Error submitting proofs to $NETWORK, retrying in 60 seconds"
+    send_slack_message "Error submitting proofs to $NETWORK, retrying in 60 seconds"
+    sleep 60
+    continue
+  fi
+
   echo "Waiting $VERIFICATION_WAIT_TIME seconds for verification"
   sleep $VERIFICATION_WAIT_TIME
 

batcher/Cargo.lock

@@ -26,9 +26,9 @@ use std::time::Duration;
 use aligned_sdk::core::constants::{
     ADDITIONAL_SUBMISSION_GAS_COST_PER_PROOF, BATCHER_SUBMISSION_BASE_GAS_COST,
     BUMP_BACKOFF_FACTOR, BUMP_MAX_RETRIES, BUMP_MAX_RETRY_DELAY, BUMP_MIN_RETRY_DELAY,
-    CONNECTION_TIMEOUT, DEFAULT_MAX_FEE_PER_PROOF, ETHEREUM_CALL_BACKOFF_FACTOR,
-    ETHEREUM_CALL_MAX_RETRIES, ETHEREUM_CALL_MAX_RETRY_DELAY, ETHEREUM_CALL_MIN_RETRY_DELAY,
-    GAS_PRICE_PERCENTAGE_MULTIPLIER, PERCENTAGE_DIVIDER,
+    CBOR_ARRAY_MAX_OVERHEAD, CONNECTION_TIMEOUT, DEFAULT_MAX_FEE_PER_PROOF,
+    ETHEREUM_CALL_BACKOFF_FACTOR, ETHEREUM_CALL_MAX_RETRIES, ETHEREUM_CALL_MAX_RETRY_DELAY,
+    ETHEREUM_CALL_MIN_RETRY_DELAY, GAS_PRICE_PERCENTAGE_MULTIPLIER, PERCENTAGE_DIVIDER,
     RESPOND_TO_TASK_FEE_LIMIT_PERCENTAGE_MULTIPLIER,
 };
 use aligned_sdk::core::types::{
@@ -115,6 +115,16 @@ impl Batcher {
         let s3_client = s3::create_client(upload_endpoint).await;
 
         let config = ConfigFromYaml::new(config_file);
+        // Ensure max_batch_bytes_size can at least hold one proof of max_proof_size,
+        // including the overhead introduced by serialization
+        assert!(
+            config.batcher.max_proof_size + CBOR_ARRAY_MAX_OVERHEAD
+                <= config.batcher.max_batch_byte_size,
+            "max_batch_bytes_size ({}) not big enough for one max_proof_size ({}) proof",
+            config.batcher.max_batch_byte_size,
+            config.batcher.max_proof_size
+        );
+
         let deployment_output =
             ContractDeploymentOutput::new(config.aligned_layer_deployment_config_file_path);
 

batcher/aligned-batcher/src/lib.rs

@@ -1,6 +1,9 @@
 use aligned_sdk::{
     communication::serialization::cbor_serialize,
-    core::types::{NoncedVerificationData, VerificationDataCommitment},
+    core::{
+        constants::CBOR_ARRAY_MAX_OVERHEAD,
+        types::{NoncedVerificationData, VerificationDataCommitment},
+    },
 };
 use ethers::types::{Address, Signature, U256};
 use priority_queue::PriorityQueue;
@@ -132,7 +135,10 @@ pub(crate) fn calculate_batch_size(batch_queue: &BatchQueue) -> Result<usize, Ba
     });
 
     if let ControlFlow::Continue(batch_size) = folded_result {
-        Ok(batch_size)
+        // We over-estimate the size of the batch by at most 8 bytes.
+        // This saves us from a scenario where we actually try to send more
+        // than the maximum allowed bytes and get rejected by operators.
+        Ok(CBOR_ARRAY_MAX_OVERHEAD + batch_size)
     } else {
         Err(BatcherError::SerializationError(String::from(
             "Could not calculate size of batch",