fix: harden graph and postgres data integrity

.github/workflows/ci.yml

@@ -6,6 +6,10 @@ on:
   pull_request:
   workflow_dispatch:
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 env:
   CARGO_TERM_COLOR: always
   RUST_BACKTRACE: 1

.github/workflows/postgres-extension-ci.yml

@@ -13,6 +13,10 @@ on:
       - '.github/workflows/postgres-extension-ci.yml'
   workflow_dispatch:
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 env:
   CARGO_TERM_COLOR: always
   RUST_BACKTRACE: 1
@@ -101,6 +105,10 @@ jobs:
         run: cargo pgrx init --pg${{ matrix.pg_version }}=/opt/homebrew/opt/postgresql@${{ matrix.pg_version }}/bin/pg_config
         working-directory: crates/ruvector-postgres
 
+      - name: Configure macOS extension linking
+        if: runner.os == 'macOS'
+        run: echo 'RUSTFLAGS=-D warnings -C link-arg=-undefined -C link-arg=dynamic_lookup' >> "$GITHUB_ENV"
+
       - name: Check code formatting
         run: cargo fmt --all -- --check
         working-directory: crates/ruvector-postgres
@@ -159,6 +167,7 @@ jobs:
   benchmark:
     name: Benchmark
     runs-on: ubuntu-latest
+    timeout-minutes: 30
     if: github.event_name == 'pull_request'
 
     steps:
@@ -185,7 +194,19 @@ jobs:
         working-directory: crates/ruvector-postgres
 
       - name: Run benchmarks
-        run: cargo bench --no-default-features --features pg17 -- --output-format bencher | tee benchmark-output.txt
+        # The quantization benches link extension entry points that PostgreSQL
+        # supplies at load time, so they cannot be executed as standalone
+        # Linux binaries. Run the self-contained Criterion targets here; the
+        # quantization paths are covered by the pgrx test jobs above.
+        run: |
+          set -o pipefail
+          cargo bench --no-default-features --features pg17 \
+            --bench distance_bench \
+            --bench e2e_bench \
+            --bench hybrid_bench \
+            --bench index_bench \
+            --bench integrity_bench \
+            -- --output-format bencher | tee benchmark-output.txt
         working-directory: crates/ruvector-postgres
 
       - name: Store benchmark result

.github/workflows/ruvector-postgres-ci.yml

@@ -173,6 +173,10 @@ jobs:
         run: cargo pgrx init --pg${{ matrix.pg_version }}=/opt/homebrew/opt/postgresql@${{ matrix.pg_version }}/bin/pg_config
         working-directory: crates/ruvector-postgres
 
+      - name: Configure macOS extension linking
+        if: runner.os == 'macOS'
+        run: echo 'RUSTFLAGS=-D warnings -C link-arg=-undefined -C link-arg=dynamic_lookup' >> "$GITHUB_ENV"
+
       - name: Build extension
         run: cargo build --no-default-features --features pg${{ matrix.pg_version }} --release
         working-directory: crates/ruvector-postgres

crates/ruvector-graph/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "ruvector-graph"
-version.workspace = true
+version = "2.3.1"
 edition.workspace = true
 rust-version.workspace = true
 license.workspace = true

crates/ruvector-graph/src/distributed/replication.rs

@@ -111,7 +111,7 @@ impl GraphReplication {
                 &format!("{}:9001", primary_node),
                 ReplicaRole::Primary,
             )
-            .map_err(|e| GraphError::ReplicationError(e))?;
+            .map_err(|e| GraphError::ReplicationError(e.to_string()))?;
 
         // Add secondary replicas
         for (idx, node) in replica_nodes.iter().enumerate() {
@@ -121,7 +121,7 @@ impl GraphReplication {
                     &format!("{}:9001", node),
                     ReplicaRole::Secondary,
                 )
-                .map_err(|e| GraphError::ReplicationError(e))?;
+                .map_err(|e| GraphError::ReplicationError(e.to_string()))?;
         }
 
         let replica_set = Arc::new(replica_set);
@@ -215,8 +215,11 @@ impl GraphReplication {
             .ok_or_else(|| GraphError::ShardError(format!("Shard {} not initialized", shard_id)))?;
 
         // Serialize operation
-        let data = bincode::encode_to_vec(&op, bincode::config::standard())
-            .map_err(|e| GraphError::SerializationError(e.to_string()))?;
+        // Graph properties contain `serde_json::Value`, whose `deserialize_any`
+        // representation is not supported by bincode 2's serde adapter. JSON
+        // preserves those dynamically typed values and remains wire-stable.
+        let _data =
+            serde_json::to_vec(&op).map_err(|e| GraphError::SerializationError(e.to_string()))?;
 
         // Append to replication log
         // Note: In production, the sync_manager would handle actual replication
@@ -404,4 +407,26 @@ mod tests {
         assert_eq!(stats.total_shards, 0);
         assert_eq!(stats.strategy, ReplicationStrategy::FullShard);
     }
+
+    #[test]
+    fn replication_operations_preserve_dynamic_json_properties() {
+        let op = ReplicationOp::AddNode(NodeData {
+            id: "node-1".to_string(),
+            properties: HashMap::from([(
+                "kind".to_string(),
+                serde_json::Value::String("test".to_string()),
+            )]),
+            labels: vec!["Example".to_string()],
+        });
+
+        let encoded = serde_json::to_vec(&op).unwrap();
+        let decoded: ReplicationOp = serde_json::from_slice(&encoded).unwrap();
+        match decoded {
+            ReplicationOp::AddNode(node) => {
+                assert_eq!(node.id, "node-1");
+                assert_eq!(node.labels, vec!["Example"]);
+            }
+            _ => panic!("unexpected replication operation"),
+        }
+    }
 }