Use hardcoded docker override

98aa9c1152

.github/workflows/release.yml

@@ -101,11 +101,11 @@ jobs:
       working-directory: src
 
     # Run tests
-    # - name: L0
-    #   run: |
-    #     ${{ matrix.devScript }} test
-    #   working-directory: src
-    #   if: matrix.runtime != 'linux-arm64' && matrix.runtime != 'linux-arm'
+    - name: L0
+      run: |
+        ${{ matrix.devScript }} test
+      working-directory: src
+      if: matrix.runtime != 'linux-arm64' && matrix.runtime != 'linux-arm'
 
     # Create runner package tar.gz/zip
     - name: Package Release

releaseNote.md

@@ -1,7 +1,11 @@
 ## Features
 
 ## Bugs
-- Fixed an issue where websockets failed to successfully close when posting log lines (#1790)
+- Fixed a crash on runner startup (#1770)
+
+## Misc
+
+- Clarified the type of step running when running job started or completed hooks (#1769)
 
 
 ## Windows x64

releaseVersion

@@ -1 +1 @@
-2.289.3
+<Update to ./src/runnerversion when creating release>

src/Misc/layoutroot/run-helper.cmd.template

@@ -0,0 +1,39 @@
+@echo off
+
+"%~dp0\bin\Runner.Listener.exe" run %*
+
+rem using `if %ERRORLEVEL% EQU N` insterad of `if ERRORLEVEL N`
+rem `if ERRORLEVEL N` means: error level is N or MORE
+  
+if %ERRORLEVEL% EQU 0 (
+  echo "Runner listener exit with 0 return code, stop the service, no retry needed."
+  exit /b 0
+)
+
+if %ERRORLEVEL% EQU 1 (
+  echo "Runner listener exit with terminated error, stop the service, no retry needed."
+  exit /b 0
+)
+
+if %ERRORLEVEL% EQU 2 (
+  echo "Runner listener exit with retryable error, re-launch runner in 5 seconds."
+  ping 127.0.0.1 -n 6 -w 1000 >NUL
+  exit /b 1
+)
+
+if %ERRORLEVEL% EQU 3 (
+  rem Sleep 5 seconds to wait for the runner update process finish
+  echo "Runner listener exit because of updating, re-launch runner in 5 seconds"
+  ping 127.0.0.1 -n 6 -w 1000 >NUL
+  exit /b 1
+)
+
+if %ERRORLEVEL% EQU 4 (
+  rem Sleep 5 seconds to wait for the ephemeral runner update process finish
+  echo "Runner listener exit because of updating, re-launch ephemeral runner in 5 seconds"
+  ping 127.0.0.1 -n 6 -w 1000 >NUL
+  exit /b 1
+)
+
+echo "Exiting after unknown error code: %ERRORLEVEL%"
+exit /b 0

src/Misc/layoutroot/run-helper.sh.template

@@ -0,0 +1,46 @@
+#!/bin/bash
+
+# Validate not sudo
+user_id=`id -u`
+if [ $user_id -eq 0 -a -z "$RUNNER_ALLOW_RUNASROOT" ]; then
+    echo "Must not run interactively with sudo"
+    exit 1
+fi
+
+# Run
+shopt -s nocasematch
+
+SOURCE="${BASH_SOURCE[0]}"
+while [ -h "$SOURCE" ]; do # resolve $SOURCE until the file is no longer a symlink
+    DIR="$( cd -P "$( dirname "$SOURCE" )" && pwd )"
+    SOURCE="$(readlink "$SOURCE")"
+    [[ $SOURCE != /* ]] && SOURCE="$DIR/$SOURCE" # if $SOURCE was a relative symlink, we need to resolve it relative to the path where the symlink file was located
+done
+DIR="$( cd -P "$( dirname "$SOURCE" )" && pwd )"
+"$DIR"/bin/Runner.Listener run $*
+
+returnCode=$?
+if [[ $returnCode == 0 ]]; then
+    echo "Runner listener exit with 0 return code, stop the service, no retry needed."
+    exit 0
+elif [[ $returnCode == 1 ]]; then
+    echo "Runner listener exit with terminated error, stop the service, no retry needed."
+    exit 0
+elif [[ $returnCode == 2 ]]; then
+    echo "Runner listener exit with retryable error, re-launch runner in 5 seconds."
+    "$DIR"/safe_sleep.sh 5
+    exit 2
+elif [[ $returnCode == 3 ]]; then
+    # Sleep 5 seconds to wait for the runner update process finish
+    echo "Runner listener exit because of updating, re-launch runner in 5 seconds"
+    "$DIR"/safe_sleep.sh 5
+    exit 2
+elif [[ $returnCode == 4 ]]; then
+    # Sleep 5 seconds to wait for the ephemeral runner update process finish
+    echo "Runner listener exit because of updating, re-launch ephemeral runner in 5 seconds"
+    "$DIR"/safe_sleep.sh 5
+    exit 2
+else
+    echo "Exiting with unknown error code: ${returnCode}"
+    exit 0
+fi

src/Misc/layoutroot/run.cmd

@@ -13,21 +13,19 @@ if defined VERBOSE_ARG (
 rem Unblock files in the root of the layout folder. E.g. .cmd files.
 powershell.exe -NoLogo -Sta -NoProfile -NonInteractive -ExecutionPolicy Unrestricted -Command "$VerbosePreference = %VERBOSE_ARG% ; Get-ChildItem -LiteralPath '%~dp0' | ForEach-Object { Write-Verbose ('Unblock: {0}' -f $_.FullName) ; $_ } | Unblock-File | Out-Null"
 
-if /i "%~1" equ "localRun" (
-    rem ********************************************************************************
-    rem Local run.
-    rem ********************************************************************************
-    "%~dp0bin\Runner.Listener.exe" %*
-) else (
-  rem ********************************************************************************
-  rem Run.
-  rem ********************************************************************************
-  "%~dp0bin\Runner.Listener.exe" run %*
 
-  rem Return code 4 means the run once runner received an update message.
-  rem Sleep 5 seconds to wait for the update process finish and run the runner again.
-  if ERRORLEVEL 4 (
-    timeout /t 5 /nobreak > NUL
-    "%~dp0bin\Runner.Listener.exe" run %*
-  )
+rem ********************************************************************************
+rem Run.
+rem ********************************************************************************
+
+:launch_helper
+copy "%~dp0run-helper.cmd.template" "%~dp0run-helper.cmd" /Y
+call "%~dp0run-helper.cmd" %*
+  
+if %ERRORLEVEL% EQU 1 (
+  echo "Restarting runner..."
+  goto :launch_helper
+) else (  
+  echo "Exiting runner..."
+  exit /b 0
 )

src/Misc/layoutroot/run.sh

@@ -1,64 +1,24 @@
 #!/bin/bash
 
-# Validate not sudo
-user_id=`id -u`
-if [ $user_id -eq 0 -a -z "$RUNNER_ALLOW_RUNASROOT" ]; then
-    echo "Must not run interactively with sudo"
-    exit 1
-fi
-
 # Change directory to the script root directory
 # https://stackoverflow.com/questions/59895/getting-the-source-directory-of-a-bash-script-from-within
 SOURCE="${BASH_SOURCE[0]}"
 while [ -h "$SOURCE" ]; do # resolve $SOURCE until the file is no longer a symlink
-  DIR="$( cd -P "$( dirname "$SOURCE" )" && pwd )"
-  SOURCE="$(readlink "$SOURCE")"
-  [[ $SOURCE != /* ]] && SOURCE="$DIR/$SOURCE" # if $SOURCE was a relative symlink, we need to resolve it relative to the path where the symlink file was located
+    DIR="$( cd -P "$( dirname "$SOURCE" )" && pwd )"
+    SOURCE="$(readlink "$SOURCE")"
+    [[ $SOURCE != /* ]] && SOURCE="$DIR/$SOURCE" # if $SOURCE was a relative symlink, we need to resolve it relative to the path where the symlink file was located
 done
 DIR="$( cd -P "$( dirname "$SOURCE" )" && pwd )"
-
-# Do not "cd $DIR". For localRun, the current directory is expected to be the repo location on disk.
-
-# Run
-shopt -s nocasematch
-if [[ "$1" == "localRun" ]]; then
-    "$DIR"/bin/Runner.Listener $*
-else
-    "$DIR"/bin/Runner.Listener run $*
-
-# Return code 3 means the run once runner received an update message.
-# Sleep 5 seconds to wait for the update process finish
+cp -f "$DIR"/run-helper.sh.template "$DIR"/run-helper.sh
+# run the helper process which keep the listener alive
+while :;
+do
+    "$DIR"/run-helper.sh $*
     returnCode=$?
-    if [[ $returnCode == 3 ]]; then
-        if [ ! -x "$(command -v sleep)" ]; then
-            if [ ! -x "$(command -v ping)" ]; then
-                COUNT="0"
-                while [[ $COUNT != 5000 ]]; do
-                    echo "SLEEP" > /dev/null
-                    COUNT=$[$COUNT+1]
-                done
-            else
-                ping -c 5 127.0.0.1 > /dev/null
-            fi
-        else
-            sleep 5
-        fi
-    elif [[ $returnCode == 4 ]]; then
-        if [ ! -x "$(command -v sleep)" ]; then
-            if [ ! -x "$(command -v ping)" ]; then
-                COUNT="0"
-                while [[ $COUNT != 5000 ]]; do
-                    echo "SLEEP" > /dev/null
-                    COUNT=$[$COUNT+1]
-                done
-            else
-                ping -c 5 127.0.0.1 > /dev/null
-            fi
-        else
-            sleep 5
-        fi
-        "$DIR"/bin/Runner.Listener run $*
+    if [[ $returnCode -eq 2 ]]; then
+        echo "Restarting runner..."
     else
-        exit $returnCode
+        echo "Exiting runner..."
+        exit 0
     fi
-fi
+done

src/Runner.Common/JobServer.cs

@@ -143,10 +143,8 @@ namespace GitHub.Runner.Common
 
         public ValueTask DisposeAsync()
         {
-            CloseWebSocket(WebSocketCloseStatus.NormalClosure, CancellationToken.None);
-
+            _websocketClient?.CloseOutputAsync(WebSocketCloseStatus.NormalClosure, "Shutdown", CancellationToken.None);
             GC.SuppressFinalize(this);
-
             return ValueTask.CompletedTask;
         }
 
@@ -250,8 +248,7 @@ namespace GitHub.Runner.Common
                         if (failedAttemptsToPostBatchedLinesByWebsocket * 100 / totalBatchedLinesAttemptedByWebsocket > _minWebsocketFailurePercentageAllowed)
                         {
                             Trace.Info($"Exhausted websocket allowed retries, we will not attempt websocket connection for this job to post lines again.");
-                            CloseWebSocket(WebSocketCloseStatus.InternalServerError, cancellationToken);
-
+                            _websocketClient?.CloseOutputAsync(WebSocketCloseStatus.InternalServerError, "Shutdown due to failures", cancellationToken);
                             // By setting it to null, we will ensure that we never try websocket path again for this job
                             _websocketClient = null;
                         }
@@ -279,19 +276,6 @@ namespace GitHub.Runner.Common
             }
         }
 
-        private void CloseWebSocket(WebSocketCloseStatus closeStatus, CancellationToken cancellationToken)
-        {
-            try
-            {
-                _websocketClient?.CloseOutputAsync(closeStatus, "Closing websocket", cancellationToken);
-            }
-            catch (Exception websocketEx)
-            {
-                // In some cases this might be okay since the websocket might be open yet, so just close and don't trace exceptions
-                Trace.Info($"Failed to close websocket gracefully {websocketEx.GetType().Name}");
-            }
-        }
-
         public Task<TaskAttachment> CreateAttachmentAsync(Guid scopeIdentifier, string hubName, Guid planId, Guid timelineId, Guid timelineRecordId, string type, string name, Stream uploadStream, CancellationToken cancellationToken)
         {
             CheckConnection();

src/Runner.Listener/Configuration/ConfigurationManager.cs

@@ -54,7 +54,7 @@ namespace GitHub.Runner.Listener.Configuration
             Trace.Info(nameof(LoadSettings));
             if (!IsConfigured())
             {
-                throw new InvalidOperationException("Not configured. Run config.(sh/cmd) to configure the runner.");
+                throw new NonRetryableException("Not configured. Run config.(sh/cmd) to configure the runner.");
             }
 
             RunnerSettings settings = _store.GetSettings();

src/Runner.Listener/Runner.cs

@@ -430,7 +430,7 @@ namespace GitHub.Runner.Listener
                                     }
 #endif
                                     var selfUpdater = HostContext.GetService<ISelfUpdater>();
-                                    selfUpdateTask = selfUpdater.SelfUpdate(runnerUpdateMessage, jobDispatcher, !runOnce && HostContext.StartupType != StartupType.Service, HostContext.RunnerShutdownToken);
+                                    selfUpdateTask = selfUpdater.SelfUpdate(runnerUpdateMessage, jobDispatcher, false, HostContext.RunnerShutdownToken);
                                     Trace.Info("Refresh message received, kick-off selfupdate background process.");
                                 }
                                 else

src/Runner.Worker/Container/DockerCommandManager.cs

@@ -11,7 +11,7 @@ using GitHub.Runner.Sdk;
 
 namespace GitHub.Runner.Worker.Container
 {
-    [ServiceLocator(Default = typeof(DockerCommandManager))]
+    [ServiceLocator(Default = typeof(DockerHookCommandManager))]
     public interface IDockerCommandManager : IRunnerService
     {
         string DockerPath { get; }
@@ -188,7 +188,7 @@ namespace GitHub.Runner.Worker.Container
             return outputStrings.FirstOrDefault();
         }
 
-        public async Task<int> DockerRun(IExecutionContext context, ContainerInfo container, EventHandler<ProcessDataReceivedEventArgs> stdoutDataReceived, EventHandler<ProcessDataReceivedEventArgs> stderrDataReceived)
+        public virtual async Task<int> DockerRun(IExecutionContext context, ContainerInfo container, EventHandler<ProcessDataReceivedEventArgs> stdoutDataReceived, EventHandler<ProcessDataReceivedEventArgs> stderrDataReceived)
         {
             IList<string> dockerOptions = new List<string>();
             // OPTIONS
@@ -258,7 +258,7 @@ namespace GitHub.Runner.Worker.Container
             return await ExecuteDockerCommandAsync(context, "run", optionsString, container.ContainerEnvironmentVariables, stdoutDataReceived, stderrDataReceived, context.CancellationToken);
         }
 
-        public async Task<int> DockerStart(IExecutionContext context, string containerId)
+        public virtual async Task<int> DockerStart(IExecutionContext context, string containerId)
         {
             return await ExecuteDockerCommandAsync(context, "start", containerId, context.CancellationToken);
         }

src/Runner.Worker/Container/DockerHookCommandManager.cs

@@ -0,0 +1,50 @@
+using System;
+using System.Collections.Generic;
+using System.IO;
+using System.Linq;
+using System.Threading.Tasks;
+using GitHub.DistributedTask.Pipelines;
+using GitHub.Runner.Common.Util;
+using GitHub.Runner.Worker.Handlers;
+
+namespace GitHub.Runner.Worker.Container
+{
+    public class DockerHookCommandManager : DockerCommandManager
+    {
+        public override async Task<int> DockerStart(IExecutionContext context, string containerId)
+        {
+            // check for env var
+            // execute script
+            
+            // Create the handler data.
+            var path = "/home/ferenc/Documents/runner/_layout/docker_run.sh";
+            var scriptDirectory = Path.GetDirectoryName(path);
+            var stepHost = HostContext.CreateService<IDefaultStepHost>();
+            var prependPath = string.Join(Path.PathSeparator.ToString(), context.Global.PrependPath.Reverse<string>());
+            Dictionary<string, string> inputs = new()
+            {
+                ["script"] = $"CONT_ID={containerId} " + "/usr/bin/bash" + " " + path,
+                // /bin/bash
+                ["shell"] = ScriptHandlerHelpers.GetDefaultShellForScript(path, Trace, prependPath)
+            };
+
+            // Create the handler
+            var handlerFactory = HostContext.GetService<IHandlerFactory>();
+            var handler = handlerFactory.Create(
+                            context,
+                            action: new ScriptReference(),
+                            stepHost,
+                            new ScriptActionExecutionData(),
+                            inputs,
+                            environment: new Dictionary<string, string>(VarUtil.EnvironmentVariableKeyComparer),
+                            context.Global.Variables,
+                            actionDirectory: scriptDirectory,
+                            localActionContainerSetupSteps: null);
+            handler.PrepareExecution(ActionRunStage.Main); // TODO: find out stage
+
+            await handler.RunAsync(ActionRunStage.Main);
+
+            return ((int?) handler.ExecutionContext.CommandResult) ?? 0;
+        }
+    }
+}

src/runnerversion

@@ -1 +1 @@
-2.289.3
+2.289.1