.gitea/workflows/ci.yml

@@ -49,6 +49,13 @@ jobs:
       - name: Build NonWindows solution
         run: dotnet build src/ZB.MOM.WW.MxGateway.NonWindows.slnx -c Release
 
+      # GitHub-hosted runners ship pwsh; the self-hosted act image does not. Install it as a
+      # .NET global tool (the SDK is already set up above) so the codegen check's `shell: pwsh` works.
+      - name: Install PowerShell (pwsh)
+        run: |
+          dotnet tool install --global PowerShell
+          echo "$HOME/.dotnet/tools" >> "$GITHUB_PATH"
+
       # IPC-01 / IPC-19 / IPC-20: descriptor set + Contracts/Generated must match the current protos.
       - name: Codegen / descriptor freshness
         shell: pwsh
@@ -93,11 +100,23 @@ jobs:
         with:
           distribution: temurin
           java-version: '17'
+      # GitHub-hosted runners ship gradle on PATH; the self-hosted act image does not, and the repo
+      # has no gradle wrapper. act also can't resolve the gradle/actions monorepo action, so install
+      # gradle directly (pinned to 9.5.1, matching the local homebrew build in CLAUDE.md/memory).
+      - name: Install Gradle 9.5.1
+        run: |
+          curl -sSL "https://services.gradle.org/distributions/gradle-9.5.1-bin.zip" -o /tmp/gradle.zip
+          sudo unzip -q -d /opt/gradle /tmp/gradle.zip
+          echo "/opt/gradle/gradle-9.5.1/bin" >> "$GITHUB_PATH"
       - name: Gradle test
         working-directory: clients/java
         run: gradle test
       - name: Revert spurious protobuf-version churn (no .proto changed)
-        run: git checkout -- clients/java/src/main/generated/main/java/mxaccess_gateway/v1/MxaccessGateway.java || true
+        # Both generated aggregates can pick up protobuf-runtime-version churn on regen; revert
+        # both so verify-clean still catches a real, uncommitted proto/codegen change elsewhere.
+        run: |
+          git checkout -- clients/java/src/main/generated/main/java/mxaccess_gateway/v1/MxaccessGateway.java || true
+          git checkout -- clients/java/src/main/generated/main/java/mxaccess_worker/v1/MxaccessWorker.java || true
       - name: Verify generated tree is clean
         run: git diff --exit-code -- clients/java/src/main/generated
 
@@ -105,8 +124,23 @@ jobs:
     # Self-hosted windev runner (10.100.0.48): the only host that can build the x86 / net48 Worker
     # and MXAccess-adjacent code. Not required for a green portable build; runs where a runner exists.
     runs-on: [self-hosted, windows]
+    # DISABLED pending a working Windows runner. act_runner v0.6.1 host-mode on Windows is broken:
+    # its hostexecutor writes each step's script to a path it then cannot resolve ("The system
+    # cannot find the path specified" / pwsh "not recognized"), independent of shell, and it cannot
+    # stage JS actions (actions/checkout dist/ missing). Windows containers are impractical for the
+    # net48/x86/MXAccess Worker. Until this is resolved, x86 Worker verification stays on the manual
+    # windev worktree process. Re-enable by setting repo variable WINDOWS_CI_ENABLED=true.
+    if: ${{ vars.WINDOWS_CI_ENABLED == 'true' }}
+    defaults:
+      run:
+        shell: cmd
     steps:
-      - uses: actions/checkout@v4
+      - name: Checkout (manual)
+        run: |
+          git init -q .
+          git remote add origin https://gitea.dohertylan.com/%GITHUB_REPOSITORY%
+          git -c protocol.version=2 fetch -q --depth 1 origin %GITHUB_SHA%
+          git checkout -q --force FETCH_HEAD
       - name: Build x86 Worker
         run: dotnet build src/ZB.MOM.WW.MxGateway.Worker/ZB.MOM.WW.MxGateway.Worker.csproj -p:Platform=x86
       - name: Worker tests
@@ -114,11 +148,21 @@ jobs:
 
   live-mxaccess:
     # Opt-in, scheduled only — needs installed MXAccess COM + live provider state. Never gates a push.
-    if: ${{ github.event_name == 'schedule' }}
+    # Also gated by WINDOWS_CI_ENABLED (see the windows job): the Windows host-mode runner is broken.
+    if: ${{ github.event_name == 'schedule' && vars.WINDOWS_CI_ENABLED == 'true' }}
     runs-on: [self-hosted, windows, mxaccess]
     env:
       MXGATEWAY_RUN_LIVE_MXACCESS_TESTS: '1'
+    # Same act host-mode constraint as the windows job above.
+    defaults:
+      run:
+        shell: cmd
     steps:
-      - uses: actions/checkout@v4
+      - name: Checkout (manual)
+        run: |
+          git init -q .
+          git remote add origin https://gitea.dohertylan.com/%GITHUB_REPOSITORY%
+          git -c protocol.version=2 fetch -q --depth 1 origin %GITHUB_SHA%
+          git checkout -q --force FETCH_HEAD
       - name: Live MXAccess smoke
         run: dotnet test src/ZB.MOM.WW.MxGateway.IntegrationTests/ZB.MOM.WW.MxGateway.IntegrationTests.csproj --filter FullyQualifiedName~WorkerLiveMxAccessSmokeTests

archreview/remediation/00-tracking.md

@@ -215,7 +215,7 @@ Full design + implementation for each row lives in the linked domain doc under i
 |---|---|:-:|:-:|---|---|---|
 | TST-01 | High | P2 | L | TST-04 | Not started | Reconnect/replay has no e2e test and no client consumer |
 | TST-02 | High | P0 | M | TST-04 | Done | Reconnect owner re-validation not implemented |
-| TST-03 | High | P1 | M | — | In review | No CI exists |
+| TST-03 | High | P1 | M | — | Done | No CI exists |
 | TST-04 | High | P2 | L | — | Not started | Session-resilience epic 16/28 tasks unfinished |
 | TST-05 | Medium | P1 | S | TST-03 | Not started | Real-worker control/COM paths verified opt-in only |
 | TST-06 | Medium | — | M | — | Not started | Dashboard live-data path untested |
@@ -265,3 +265,4 @@ Findings the review flagged as one coordinated design pass — sequence them tog
 | 2026-07-09 | P1 Wave 2a (security). SEC-01/04/06 → `Done` (config path-rooting + production validator guards; Server build clean, validator+hygiene tests 53/53). SEC-10 → `Done`: the shared `ZB.MOM.WW.Auth.ApiKeys` gained optional `ExpiresUtc` (expired keys rejected, auth DB auto-migrates to schema v3) via a concurrent HistorianGateway-remediation session's "G-2"; this repo consumes it by bumping the four `Auth.*` refs 0.1.2→0.1.4 (commit 197731a). Remaining SEC-10 polish (`apikey create --expires` + dashboard staleness badge) tracked as a small follow-up. |
 | 2026-07-09 | P1 Wave 1 (CI + codegen freshness + Rust buildability) via parallel agents. IPC-01/09/19/20, CLI-02 → `Done`; TST-03 → `In review` (CI pipeline authored + YAML/layout-validated, but not yet executed on a Gitea runner — proven on first push). Verified on macOS: NonWindows build clean, `ClientProtoInputTests` 5/5, `publish-client-proto-inputs.ps1 -Check` exit 0, `cargo package` (no `--no-verify`) compiles standalone. Added a vendored-Rust-proto drift guard (Check 3) to `check-codegen.ps1` closing the CLI-02 static-copy risk. IPC-09 script guards not executed end-to-end (need pinned python/JRE toolchains); logic is PATH-resolution + version assertion. |
 | 2026-07-09 | WRK-01 → `Done`. Verified on Windows host (windev) via an isolated `origin/main` worktree: worker builds x86 clean, `StaRuntimeTests`+`WorkerPipeSessionTests` 33/33 pass. Fixed an `xUnit1030` build error (the new worker test used `.ConfigureAwait(false)` in `[Fact]` bodies) that the macOS tree could not surface. Also ran GWC-01's Windows-only `WorkerClientTests` on windev: 18/18 pass (incl. `ReadEventsAsync_SecondEnumerator_Throws`). All 8 P0 findings now `Done`. Not yet committed. |
+| 2026-07-10 | **TST-03 → `Done`: CI is live and green** on branch `fix/ci-selfhosted-tooling`. The pipeline was authored (P1) but had never executed. Root cause it never ran: the co-located `gitea-runner` on the docker host (`10.100.0.35`) spawned job containers on an isolated network (`container.network: ""`) that could not resolve Gitea's internal clone URL `http://gitea:3000`; one-line host fix `container.network: "traefik"` + `docker restart gitea-runner`. The self-hosted `catthehacker` act image also lacks tooling GitHub-hosted runners preinstall — ci.yml now installs pwsh (dotnet global tool, for `check-codegen.ps1`) and Gradle 9.5.1 directly (act can't resolve the `gradle/actions` monorepo action; no gradle wrapper in repo). Driving to green surfaced and fixed **five real latent defects** (TST-03 doing its job): `check-codegen.ps1` `.Trim()`-on-`$null` on a clean tree; **stale vendored rust proto** (`clients/rust/protos/mxaccess_worker.proto` missing canonical `max_frame_bytes`); `OrphanWorkerTerminatorTests` hard-coded `C:\` path failing Linux `Path.GetFullPath` (0 kills); **stale java generated** `MxaccessWorker.java` (missing `max_frame_bytes`, regenerated); a sync python test building a `grpc.aio.Channel` with no current event loop on py3.12 (autouse conftest fixture). Result: `portable` **success** (NonWindows build + codegen freshness + 808/808 gateway tests + .NET/Go/Rust/Python clients) and `java` **success**. `windows` + `live-mxaccess` jobs remain `queued` pending a self-hosted **windev** runner (`10.100.0.48`) with those labels — separate follow-up, does not gate portable/java. |

clients/java/src/main/generated/main/java/mxaccess_worker/v1/MxaccessWorker.java

@@ -3789,6 +3789,20 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
      */
     com.google.protobuf.ByteString
         getGatewayVersionBytes();
+
+    /**
+     * <pre>
+     * Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+     * configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+     * instead of a hard-coded default; 0 (an older gateway that never set the field) means
+     * "use the worker's built-in default". Sits above the public gRPC cap by an
+     * envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+     * </pre>
+     *
+     * <code>uint32 max_frame_bytes = 4;</code>
+     * @return The maxFrameBytes.
+     */
+    int getMaxFrameBytes();
   }
   /**
    * Protobuf type {@code mxaccess_worker.v1.GatewayHello}
@@ -3918,6 +3932,25 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
       }
     }
 
+    public static final int MAX_FRAME_BYTES_FIELD_NUMBER = 4;
+    private int maxFrameBytes_ = 0;
+    /**
+     * <pre>
+     * Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+     * configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+     * instead of a hard-coded default; 0 (an older gateway that never set the field) means
+     * "use the worker's built-in default". Sits above the public gRPC cap by an
+     * envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+     * </pre>
+     *
+     * <code>uint32 max_frame_bytes = 4;</code>
+     * @return The maxFrameBytes.
+     */
+    @java.lang.Override
+    public int getMaxFrameBytes() {
+      return maxFrameBytes_;
+    }
+
     private byte memoizedIsInitialized = -1;
     @java.lang.Override
     public final boolean isInitialized() {
@@ -3941,6 +3974,9 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
       if (!com.google.protobuf.GeneratedMessage.isStringEmpty(gatewayVersion_)) {
         com.google.protobuf.GeneratedMessage.writeString(output, 3, gatewayVersion_);
       }
+      if (maxFrameBytes_ != 0) {
+        output.writeUInt32(4, maxFrameBytes_);
+      }
       getUnknownFields().writeTo(output);
     }
 
@@ -3960,6 +3996,10 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
       if (!com.google.protobuf.GeneratedMessage.isStringEmpty(gatewayVersion_)) {
         size += com.google.protobuf.GeneratedMessage.computeStringSize(3, gatewayVersion_);
       }
+      if (maxFrameBytes_ != 0) {
+        size += com.google.protobuf.CodedOutputStream
+          .computeUInt32Size(4, maxFrameBytes_);
+      }
       size += getUnknownFields().getSerializedSize();
       memoizedSize = size;
       return size;
@@ -3981,6 +4021,8 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
           .equals(other.getNonce())) return false;
       if (!getGatewayVersion()
           .equals(other.getGatewayVersion())) return false;
+      if (getMaxFrameBytes()
+          != other.getMaxFrameBytes()) return false;
       if (!getUnknownFields().equals(other.getUnknownFields())) return false;
       return true;
     }
@@ -3998,6 +4040,8 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
       hash = (53 * hash) + getNonce().hashCode();
       hash = (37 * hash) + GATEWAY_VERSION_FIELD_NUMBER;
       hash = (53 * hash) + getGatewayVersion().hashCode();
+      hash = (37 * hash) + MAX_FRAME_BYTES_FIELD_NUMBER;
+      hash = (53 * hash) + getMaxFrameBytes();
       hash = (29 * hash) + getUnknownFields().hashCode();
       memoizedHashCode = hash;
       return hash;
@@ -4132,6 +4176,7 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
         supportedProtocolVersion_ = 0;
         nonce_ = "";
         gatewayVersion_ = "";
+        maxFrameBytes_ = 0;
         return this;
       }
 
@@ -4174,6 +4219,9 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
         if (((from_bitField0_ & 0x00000004) != 0)) {
           result.gatewayVersion_ = gatewayVersion_;
         }
+        if (((from_bitField0_ & 0x00000008) != 0)) {
+          result.maxFrameBytes_ = maxFrameBytes_;
+        }
       }
 
       @java.lang.Override
@@ -4201,6 +4249,9 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
           bitField0_ |= 0x00000004;
           onChanged();
         }
+        if (other.getMaxFrameBytes() != 0) {
+          setMaxFrameBytes(other.getMaxFrameBytes());
+        }
         this.mergeUnknownFields(other.getUnknownFields());
         onChanged();
         return this;
@@ -4242,6 +4293,11 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
                 bitField0_ |= 0x00000004;
                 break;
               } // case 26
+              case 32: {
+                maxFrameBytes_ = input.readUInt32();
+                bitField0_ |= 0x00000008;
+                break;
+              } // case 32
               default: {
                 if (!super.parseUnknownField(input, extensionRegistry, tag)) {
                   done = true; // was an endgroup tag
@@ -4435,6 +4491,62 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
         return this;
       }
 
+      private int maxFrameBytes_ ;
+      /**
+       * <pre>
+       * Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+       * configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+       * instead of a hard-coded default; 0 (an older gateway that never set the field) means
+       * "use the worker's built-in default". Sits above the public gRPC cap by an
+       * envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+       * </pre>
+       *
+       * <code>uint32 max_frame_bytes = 4;</code>
+       * @return The maxFrameBytes.
+       */
+      @java.lang.Override
+      public int getMaxFrameBytes() {
+        return maxFrameBytes_;
+      }
+      /**
+       * <pre>
+       * Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+       * configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+       * instead of a hard-coded default; 0 (an older gateway that never set the field) means
+       * "use the worker's built-in default". Sits above the public gRPC cap by an
+       * envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+       * </pre>
+       *
+       * <code>uint32 max_frame_bytes = 4;</code>
+       * @param value The maxFrameBytes to set.
+       * @return This builder for chaining.
+       */
+      public Builder setMaxFrameBytes(int value) {
+
+        maxFrameBytes_ = value;
+        bitField0_ |= 0x00000008;
+        onChanged();
+        return this;
+      }
+      /**
+       * <pre>
+       * Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+       * configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+       * instead of a hard-coded default; 0 (an older gateway that never set the field) means
+       * "use the worker's built-in default". Sits above the public gRPC cap by an
+       * envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+       * </pre>
+       *
+       * <code>uint32 max_frame_bytes = 4;</code>
+       * @return This builder for chaining.
+       */
+      public Builder clearMaxFrameBytes() {
+        bitField0_ = (bitField0_ & ~0x00000008);
+        maxFrameBytes_ = 0;
+        onChanged();
+        return this;
+      }
+
       // @@protoc_insertion_point(builder_scope:mxaccess_worker.v1.GatewayHello)
     }
 
@@ -12552,64 +12664,65 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
       "rker_heartbeat\030\023 \001(\0132#.mxaccess_worker.v" +
       "1.WorkerHeartbeatH\000\0227\n\014worker_fault\030\024 \001(" +
       "\0132\037.mxaccess_worker.v1.WorkerFaultH\000B\006\n\004" +
-      "body\"Z\n\014GatewayHello\022\"\n\032supported_protoc" +
+      "body\"s\n\014GatewayHello\022\"\n\032supported_protoc" +
       "ol_version\030\001 \001(\r\022\r\n\005nonce\030\002 \001(\t\022\027\n\017gatew" +
-      "ay_version\030\003 \001(\t\"i\n\013WorkerHello\022\030\n\020proto" +
+      "ay_version\030\003 \001(\t\022\027\n\017max_frame_bytes\030\004 \001(" +
-      "col_version\030\001 \001(\r\022\r\n\005nonce\030\002 \001(\t\022\031\n\021work" +
+      "\r\"i\n\013WorkerHello\022\030\n\020protocol_version\030\001 \001" +
-      "er_process_id\030\003 \001(\005\022\026\n\016worker_version\030\004 " +
+      "(\r\022\r\n\005nonce\030\002 \001(\t\022\031\n\021worker_process_id\030\003" +
-      "\001(\t\"\216\001\n\013WorkerReady\022\031\n\021worker_process_id" +
+      " \001(\005\022\026\n\016worker_version\030\004 \001(\t\"\216\001\n\013WorkerR" +
-      "\030\001 \001(\005\022\027\n\017mxaccess_progid\030\002 \001(\t\022\026\n\016mxacc" +
+      "eady\022\031\n\021worker_process_id\030\001 \001(\005\022\027\n\017mxacc" +
-      "ess_clsid\030\003 \001(\t\0223\n\017ready_timestamp\030\004 \001(\013" +
+      "ess_progid\030\002 \001(\t\022\026\n\016mxaccess_clsid\030\003 \001(\t" +
-      "2\032.google.protobuf.Timestamp\"w\n\rWorkerCo" +
+      "\0223\n\017ready_timestamp\030\004 \001(\0132\032.google.proto" +
-      "mmand\022/\n\007command\030\001 \001(\0132\036.mxaccess_gatewa" +
+      "buf.Timestamp\"w\n\rWorkerCommand\022/\n\007comman" +
-      "y.v1.MxCommand\0225\n\021enqueue_timestamp\030\002 \001(" +
+      "d\030\001 \001(\0132\036.mxaccess_gateway.v1.MxCommand\022" +
-      "\0132\032.google.protobuf.Timestamp\"\201\001\n\022Worker" +
+      "5\n\021enqueue_timestamp\030\002 \001(\0132\032.google.prot" +
-      "CommandReply\0222\n\005reply\030\001 \001(\0132#.mxaccess_g" +
+      "obuf.Timestamp\"\201\001\n\022WorkerCommandReply\0222\n" +
-      "ateway.v1.MxCommandReply\0227\n\023completed_ti" +
+      "\005reply\030\001 \001(\0132#.mxaccess_gateway.v1.MxCom" +
-      "mestamp\030\002 \001(\0132\032.google.protobuf.Timestam" +
+      "mandReply\0227\n\023completed_timestamp\030\002 \001(\0132\032" +
-      "p\"\036\n\014WorkerCancel\022\016\n\006reason\030\001 \001(\t\"Q\n\016Wor" +
+      ".google.protobuf.Timestamp\"\036\n\014WorkerCanc" +
-      "kerShutdown\022/\n\014grace_period\030\001 \001(\0132\031.goog" +
+      "el\022\016\n\006reason\030\001 \001(\t\"Q\n\016WorkerShutdown\022/\n\014" +
-      "le.protobuf.Duration\022\016\n\006reason\030\002 \001(\t\"H\n\021" +
+      "grace_period\030\001 \001(\0132\031.google.protobuf.Dur" +
-      "WorkerShutdownAck\0223\n\006status\030\001 \001(\0132#.mxac" +
+      "ation\022\016\n\006reason\030\002 \001(\t\"H\n\021WorkerShutdownA" +
-      "cess_gateway.v1.ProtocolStatus\":\n\013Worker" +
+      "ck\0223\n\006status\030\001 \001(\0132#.mxaccess_gateway.v1" +
-      "Event\022+\n\005event\030\001 \001(\0132\034.mxaccess_gateway." +
+      ".ProtocolStatus\":\n\013WorkerEvent\022+\n\005event\030" +
-      "v1.MxEvent\"\245\002\n\017WorkerHeartbeat\022\031\n\021worker" +
+      "\001 \001(\0132\034.mxaccess_gateway.v1.MxEvent\"\245\002\n\017" +
-      "_process_id\030\001 \001(\005\022.\n\005state\030\002 \001(\0162\037.mxacc" +
+      "WorkerHeartbeat\022\031\n\021worker_process_id\030\001 \001" +
-      "ess_worker.v1.WorkerState\022?\n\033last_sta_ac" +
+      "(\005\022.\n\005state\030\002 \001(\0162\037.mxaccess_worker.v1.W" +
-      "tivity_timestamp\030\003 \001(\0132\032.google.protobuf" +
+      "orkerState\022?\n\033last_sta_activity_timestam" +
-      ".Timestamp\022\035\n\025pending_command_count\030\004 \001(" +
+      "p\030\003 \001(\0132\032.google.protobuf.Timestamp\022\035\n\025p" +
-      "\r\022\"\n\032outbound_event_queue_depth\030\005 \001(\r\022\033\n" +
+      "ending_command_count\030\004 \001(\r\022\"\n\032outbound_e" +
-      "\023last_event_sequence\030\006 \001(\004\022&\n\036current_co" +
+      "vent_queue_depth\030\005 \001(\r\022\033\n\023last_event_seq" +
-      "mmand_correlation_id\030\007 \001(\t\"\364\001\n\013WorkerFau" +
+      "uence\030\006 \001(\004\022&\n\036current_command_correlati" +
-      "lt\0229\n\010category\030\001 \001(\0162\'.mxaccess_worker.v" +
+      "on_id\030\007 \001(\t\"\364\001\n\013WorkerFault\0229\n\010category\030" +
-      "1.WorkerFaultCategory\022\026\n\016command_method\030" +
+      "\001 \001(\0162\'.mxaccess_worker.v1.WorkerFaultCa" +
-      "\002 \001(\t\022\024\n\007hresult\030\003 \001(\005H\000\210\001\001\022\026\n\016exception" +
+      "tegory\022\026\n\016command_method\030\002 \001(\t\022\024\n\007hresul" +
-      "_type\030\004 \001(\t\022\032\n\022diagnostic_message\030\005 \001(\t\022" +
+      "t\030\003 \001(\005H\000\210\001\001\022\026\n\016exception_type\030\004 \001(\t\022\032\n\022" +
-      "<\n\017protocol_status\030\006 \001(\0132#.mxaccess_gate" +
+      "diagnostic_message\030\005 \001(\t\022<\n\017protocol_sta" +
-      "way.v1.ProtocolStatusB\n\n\010_hresult*\227\002\n\013Wo" +
+      "tus\030\006 \001(\0132#.mxaccess_gateway.v1.Protocol" +
-      "rkerState\022\034\n\030WORKER_STATE_UNSPECIFIED\020\000\022" +
+      "StatusB\n\n\010_hresult*\227\002\n\013WorkerState\022\034\n\030WO" +
-      "\031\n\025WORKER_STATE_STARTING\020\001\022\034\n\030WORKER_STA" +
+      "RKER_STATE_UNSPECIFIED\020\000\022\031\n\025WORKER_STATE" +
-      "TE_HANDSHAKING\020\002\022!\n\035WORKER_STATE_INITIAL" +
+      "_STARTING\020\001\022\034\n\030WORKER_STATE_HANDSHAKING\020" +
-      "IZING_STA\020\003\022\026\n\022WORKER_STATE_READY\020\004\022\"\n\036W" +
+      "\002\022!\n\035WORKER_STATE_INITIALIZING_STA\020\003\022\026\n\022" +
-      "ORKER_STATE_EXECUTING_COMMAND\020\005\022\036\n\032WORKE" +
+      "WORKER_STATE_READY\020\004\022\"\n\036WORKER_STATE_EXE" +
-      "R_STATE_SHUTTING_DOWN\020\006\022\030\n\024WORKER_STATE_" +
+      "CUTING_COMMAND\020\005\022\036\n\032WORKER_STATE_SHUTTIN" +
-      "STOPPED\020\007\022\030\n\024WORKER_STATE_FAULTED\020\010*\307\004\n\023" +
+      "G_DOWN\020\006\022\030\n\024WORKER_STATE_STOPPED\020\007\022\030\n\024WO" +
-      "WorkerFaultCategory\022%\n!WORKER_FAULT_CATE" +
+      "RKER_STATE_FAULTED\020\010*\307\004\n\023WorkerFaultCate" +
-      "GORY_UNSPECIFIED\020\000\022+\n\'WORKER_FAULT_CATEG" +
+      "gory\022%\n!WORKER_FAULT_CATEGORY_UNSPECIFIE" +
-      "ORY_INVALID_ARGUMENTS\020\001\0227\n3WORKER_FAULT_" +
+      "D\020\000\022+\n\'WORKER_FAULT_CATEGORY_INVALID_ARG" +
-      "CATEGORY_GATEWAY_AUTHENTICATION_FAILED\020\002" +
+      "UMENTS\020\001\0227\n3WORKER_FAULT_CATEGORY_GATEWA" +
-      "\022+\n\'WORKER_FAULT_CATEGORY_PROTOCOL_MISMA" +
+      "Y_AUTHENTICATION_FAILED\020\002\022+\n\'WORKER_FAUL" +
-      "TCH\020\003\022,\n(WORKER_FAULT_CATEGORY_PROTOCOL_" +
+      "T_CATEGORY_PROTOCOL_MISMATCH\020\003\022,\n(WORKER" +
-      "VIOLATION\020\004\022+\n\'WORKER_FAULT_CATEGORY_PIP" +
+      "_FAULT_CATEGORY_PROTOCOL_VIOLATION\020\004\022+\n\'" +
-      "E_DISCONNECTED\020\005\0222\n.WORKER_FAULT_CATEGOR" +
+      "WORKER_FAULT_CATEGORY_PIPE_DISCONNECTED\020" +
-      "Y_MXACCESS_CREATION_FAILED\020\006\0221\n-WORKER_F" +
+      "\005\0222\n.WORKER_FAULT_CATEGORY_MXACCESS_CREA" +
-      "AULT_CATEGORY_MXACCESS_COMMAND_FAILED\020\007\022" +
+      "TION_FAILED\020\006\0221\n-WORKER_FAULT_CATEGORY_M" +
-      ":\n6WORKER_FAULT_CATEGORY_MXACCESS_EVENT_" +
+      "XACCESS_COMMAND_FAILED\020\007\022:\n6WORKER_FAULT" +
-      "CONVERSION_FAILED\020\010\022\"\n\036WORKER_FAULT_CATE" +
+      "_CATEGORY_MXACCESS_EVENT_CONVERSION_FAIL" +
-      "GORY_STA_HUNG\020\t\022(\n$WORKER_FAULT_CATEGORY" +
+      "ED\020\010\022\"\n\036WORKER_FAULT_CATEGORY_STA_HUNG\020\t" +
-      "_QUEUE_OVERFLOW\020\n\022*\n&WORKER_FAULT_CATEGO" +
+      "\022(\n$WORKER_FAULT_CATEGORY_QUEUE_OVERFLOW" +
-      "RY_SHUTDOWN_TIMEOUT\020\013B&\252\002#ZB.MOM.WW.MxGa" +
+      "\020\n\022*\n&WORKER_FAULT_CATEGORY_SHUTDOWN_TIM" +
-      "teway.Contracts.Protob\006proto3"
+      "EOUT\020\013B&\252\002#ZB.MOM.WW.MxGateway.Contracts" +
+      ".Protob\006proto3"
     };
     descriptor = com.google.protobuf.Descriptors.FileDescriptor
       .internalBuildGeneratedFileFrom(descriptorData,
@@ -12629,7 +12742,7 @@ public final class MxaccessWorker extends com.google.protobuf.GeneratedFile {
     internal_static_mxaccess_worker_v1_GatewayHello_fieldAccessorTable = new
       com.google.protobuf.GeneratedMessage.FieldAccessorTable(
         internal_static_mxaccess_worker_v1_GatewayHello_descriptor,
-        new java.lang.String[] { "SupportedProtocolVersion", "Nonce", "GatewayVersion", });
+        new java.lang.String[] { "SupportedProtocolVersion", "Nonce", "GatewayVersion", "MaxFrameBytes", });
     internal_static_mxaccess_worker_v1_WorkerHello_descriptor =
       getDescriptor().getMessageType(2);
     internal_static_mxaccess_worker_v1_WorkerHello_fieldAccessorTable = new

clients/python/tests/conftest.py

@@ -0,0 +1,27 @@
+"""Shared pytest fixtures for the client test suite.
+
+pytest-asyncio 1.x no longer installs a current event loop for *synchronous*
+tests, and it clears the loop after each ``async`` test. Some synchronous tests
+build a ``grpc.aio.Channel`` (via ``create_channel``), which calls
+``asyncio.get_event_loop()`` and therefore raises
+``RuntimeError: There is no current event loop`` on Python 3.12 when a prior
+async test has left the policy with no current loop. Ensure every test starts
+with a usable current loop so channel construction in a sync test is
+order-independent.
+"""
+
+from __future__ import annotations
+
+import asyncio
+from collections.abc import Iterator
+
+import pytest
+
+
+@pytest.fixture(autouse=True)
+def _ensure_current_event_loop() -> Iterator[None]:
+    try:
+        asyncio.get_event_loop()
+    except RuntimeError:
+        asyncio.set_event_loop(asyncio.new_event_loop())
+    yield

clients/rust/protos/mxaccess_worker.proto

@@ -42,6 +42,12 @@ message GatewayHello {
   uint32 supported_protocol_version = 1;
   string nonce = 2;
   string gateway_version = 3;
+  // Maximum worker-frame payload size, in bytes, negotiated by the gateway from its
+  // configured pipe limit. The worker adopts this as its frame-protocol MaxMessageBytes
+  // instead of a hard-coded default; 0 (an older gateway that never set the field) means
+  // "use the worker's built-in default". Sits above the public gRPC cap by an
+  // envelope-overhead margin so an accepted gRPC payload always fits one worker frame.
+  uint32 max_frame_bytes = 4;
 }
 
 message WorkerHello {

scripts/check-codegen.ps1

@@ -52,7 +52,9 @@ try {
         $failures.Add('Contracts project failed to build during codegen check.')
     }
 
-    $diff = (& git -C $repoRoot status --porcelain -- 'src/ZB.MOM.WW.MxGateway.Contracts/Generated').Trim()
+    # Pipe through Out-String so a clean tree (git emits nothing -> $null) does not throw
+    # "cannot call a method on a null-valued expression"; Out-String yields '' for no output.
+    $diff = (& git -C $repoRoot status --porcelain -- 'src/ZB.MOM.WW.MxGateway.Contracts/Generated' | Out-String).Trim()
     if (-not [string]::IsNullOrEmpty($diff)) {
         Write-Host $diff
         $failures.Add('Contracts/Generated differs from a fresh regeneration. Run `dotnet build` on the contracts project and commit Generated/ (required for the net48 worker build).')

src/ZB.MOM.WW.MxGateway.Tests/Gateway/Workers/OrphanWorkerTerminatorTests.cs

@@ -13,7 +13,19 @@ namespace ZB.MOM.WW.MxGateway.Tests.Gateway.Workers;
 /// </summary>
 public sealed class OrphanWorkerTerminatorTests
 {
-    private const string WorkerExecutablePath = @"C:\app\src\ZB.MOM.WW.MxGateway.Worker\bin\x86\Release\ZB.MOM.WW.MxGateway.Worker.exe";
+    // An OS-appropriate, already-normalized absolute path so Path.GetFullPath (used by the
+    // terminator to resolve the configured path) is idempotent and the exact-match compare holds
+    // on every platform. A hard-coded Windows literal (C:\...) is not rooted on Linux, so
+    // Path.GetFullPath there prepends the cwd and the match fails — the terminator is a
+    // Windows-only feature in production, but its matching logic must still be testable on the
+    // Linux CI runner.
+    private static readonly string WorkerExecutablePath = Path.GetFullPath(Path.Combine(
+        Path.GetTempPath(),
+        "MxGatewayOrphanTests",
+        "bin",
+        "x86",
+        "Release",
+        "ZB.MOM.WW.MxGateway.Worker.exe"));
 
     /// <summary>Verifies that orphan worker processes matching the configured executable path are killed.</summary>
     [Fact]
@@ -59,7 +71,10 @@ public sealed class OrphanWorkerTerminatorTests
         // not our worker and must be left alone.
         FakeProcessInspector inspector = new(
         [
-            new RunningProcessInfo(301, @"C:\other\place\ZB.MOM.WW.MxGateway.Worker.exe"),
+            new RunningProcessInfo(301, Path.GetFullPath(Path.Combine(
+                Path.GetTempPath(),
+                "MxGatewayOrphanTests-other",
+                "ZB.MOM.WW.MxGateway.Worker.exe"))),
         ]);
         OrphanWorkerTerminator terminator = CreateTerminator(inspector);