feat: buffer connection logs and flush them on failure

CHANGELOG.md

@@ -5,10 +5,12 @@
 ### Added
 
 - include a Coder support bundle for the selected workspace and agent only when using the environment's _Collect logs_
-  action in Toolbox
+  action in Toolbox; requires Coder server and CLI 2.10.0 or newer. Toolbox logs and eligible local SSH logs remain
+  available when a Coder support bundle cannot be generated
 - include active and recent local CLI SSH logs in workspace log collection. For CLI 2.4.0 and newer the plugin now
   passes an explicit `--log-dir` to the SSH ProxyCommand (the configured directory, `CODER_SSH_LOG_DIR`, or
   `ssh-logs/` in the deployment's plugin data directory) and collects and cleans up logs from that same directory
+- add a flight recorder that keeps recent plugin messages below Toolbox's log level in memory for connection diagnostics
 
 ### Changed
 
@@ -18,8 +20,7 @@
 - remove stored credentials on explicit logout, and remove old plaintext session files after successful keyring login
 - bound credential and header commands to 60 seconds while preserving unbounded workspace starts and complete progress
   lines
-
-- upgraded the Toolbox plugin API, dropping support for Toolbox versions older than 3.7.2
+- upgraded the Toolbox plugin API, dropping support for Toolbox versions older than 3.9.0
 - display live workspace start output from the Coder CLI and poll build status and provisioner output while updating,
   restarting, or stopping workspaces
 

gradle/libs.versions.toml

@@ -1,5 +1,5 @@
 [versions]
-toolbox-plugin-api = "1.13.87111"
+toolbox-plugin-api = "1.15.88839"
 kotlin = "2.3.10"
 coroutines = "1.10.2"
 serialization = "1.9.0"

src/main/kotlin/com/coder/toolbox/CoderRemoteEnvironment.kt

@@ -162,7 +162,6 @@ class CoderRemoteEnvironment(
 
     private fun refreshAvailableActions() {
         val actions = mutableListOf<ActionDescription>()
-        context.logger.debug("Refreshing available actions for workspace $id with status: $environmentStatus")
         if (environmentStatus.canStop() && agent != null) {
             actions.add(Action(context, "Open web terminal") {
                 context.logger.debug("Launching web terminal for $id...")
@@ -240,6 +239,7 @@ class CoderRemoteEnvironment(
                     Action(context, "Update and restart") {
                         withProgress("Updating and restarting workspace…") {
                             context.logger.debug(currentSessionId(), "Updating and re-starting $id...")
+                            context.logger.flush("workspace_restarting $id", currentSessionId())
                             val build = client.updateWorkspace(workspace)
                             applyWorkspaceSnapshot(workspace.copy(latestBuild = build), agent)
                             workspaceRefreshTrigger.trySend(true)
@@ -252,6 +252,7 @@ class CoderRemoteEnvironment(
                     withProgress("Stopping workspace…") {
                         tryStopSshConnection()
                         context.logger.debug(currentSessionId(), "Stopping $id...")
+                        context.logger.flush("workspace_stopping $id", currentSessionId())
                         val build = client.stopWorkspace(workspace)
                         applyWorkspaceSnapshot(workspace.copy(latestBuild = build), agent)
                         workspaceRefreshTrigger.trySend(true)
@@ -279,6 +280,7 @@ class CoderRemoteEnvironment(
                 if (confirmation == workspace.name) {
                     withProgress("Deleting workspace…") {
                         context.logger.debug(currentSessionId(), "Deleting $id...")
+                        context.logger.flush("workspace_deleting $id", currentSessionId())
                         deleteWorkspace()
                     }
                 }
@@ -301,6 +303,7 @@ class CoderRemoteEnvironment(
                     "The SSH connection to workspace $name could not be dropped in time, " +
                             "going to stop the workspace while the SSH connection is live"
                 context.logger.warn(currentSessionId(), message)
+                context.logger.flush("ssh_disconnect_timeout $id", currentSessionId())
             }
         }
     }
@@ -386,6 +389,7 @@ class CoderRemoteEnvironment(
         }
         removedSessionId?.let { sessionId ->
             context.logger.info(sessionId, "Removed Toolbox SSH session for $id")
+            context.logger.flush("ssh_session_removed $id", sessionId)
         }
     }
 
@@ -412,6 +416,7 @@ class CoderRemoteEnvironment(
             "Toolbox is disconnecting SSH from $id $disconnectKind.$stateInference " +
                     "Latest known Coder state: $latestCoderState",
         )
+        context.logger.flush("ssh_disconnected $id", sessionId)
         context.logger.info(sessionId, "Stopping the network metrics poll job for $id")
         pollJob?.cancel()
         pollJob = null
@@ -425,6 +430,7 @@ class CoderRemoteEnvironment(
                 SessionIdRegistry.removeSession(workspace.name, it.name)
             }?.let {
                 context.logger.info(it, "Removed Toolbox SSH session for $id after manual disconnect")
+                context.logger.flush("ssh_session_removed_after_manual_disconnect $id", it)
             }
         }
     }
@@ -479,6 +485,7 @@ class CoderRemoteEnvironment(
         val previousEnvironmentStatus = environmentStatus
         val previousWorkspace = workspace
         val previousAgent = agent
+        val wasReachable = state.value.isReachable
         val sessionId = currentSessionId()
 
         environmentStatus = newState
@@ -499,6 +506,9 @@ class CoderRemoteEnvironment(
                     "agent lifecycle state: ${previousAgent?.lifecycleState} -> ${agent?.lifecycleState}, " +
                     "login before ready: ${previousAgent?.loginBeforeReady} -> ${agent?.loginBeforeReady}"
         context.logger.info(sessionId, message)
+        if (wasReachable && !state.value.isReachable) {
+            context.logger.flush("environment_status_changed $id", sessionId)
+        }
     }
 
     /**

src/main/kotlin/com/coder/toolbox/CoderRemoteProvider.kt

@@ -195,6 +195,7 @@ class CoderRemoteProvider(
                                 "Your Coder session has expired. Please re-authenticate and try again.",
                                 ex
                             )
+                            context.logger.flush("coder_session_expired")
                             break
                         }
                         context.logger.error(sessionIds, ex, "workspace polling error encountered")
@@ -359,6 +360,7 @@ class CoderRemoteProvider(
         val activeClient = client
         val activeCli = cli
         context.logger.info(sessionIds, "Logging out ${activeClient?.me?.username}...")
+        context.logger.flush("logout")
         try {
             pollJob?.cancelAndJoin()
             activeClient?.close()
@@ -376,6 +378,7 @@ class CoderRemoteProvider(
                         "Sign in and try logging out again to remove it.",
                 ex,
             )
+            context.logger.flush("cli_logout_failed")
         } finally {
             close()
         }
@@ -422,6 +425,7 @@ class CoderRemoteProvider(
         accountDropdownField.visibility.update { false }
         router.clear()
         context.logger.info(sessionIds, "Coder plugin is now closed")
+        context.logger.flush("provider_closed")
     }
 
     private fun softClose() {

src/main/kotlin/com/coder/toolbox/CoderToolboxContext.kt

@@ -34,7 +34,9 @@ data class CoderToolboxContext(
     val secrets: CoderSecretsStore,
     val proxySettings: ToolboxProxySettings,
 ) {
-    val logger: CoderLogger = CoderLogger(underlyingLogger, ui, cs, i18n)
+    val logger: CoderLogger =
+        underlyingLogger as? CoderLogger ?: CoderLogger(underlyingLogger, ui, cs, i18n, settingsStore.readOnly())
+
     val connectionMonitoringService: ConnectionMonitoringService = ConnectionMonitoringService(this)
 
     /**

src/main/kotlin/com/coder/toolbox/CoderToolboxExtension.kt

@@ -1,5 +1,6 @@
 package com.coder.toolbox
 
+import com.coder.toolbox.diagnostics.CoderLogger
 import com.coder.toolbox.settings.Environment
 import com.coder.toolbox.store.CoderSecretsStore
 import com.coder.toolbox.store.CoderSettingsStore
@@ -28,9 +29,12 @@ class CoderToolboxExtension : RemoteDevExtension {
     // All services must be passed in here and threaded as necessary.
     override fun createRemoteProviderPluginInstance(serviceLocator: ServiceLocator): RemoteProvider {
         val ui = serviceLocator.getService<ToolboxUi>()
-        val logger = serviceLocator.getService(Logger::class.java)
         val cs = serviceLocator.getService<CoroutineScope>()
         val i18n = serviceLocator.getService<LocalizableStringFactory>()
+        val delegate = serviceLocator.getService<Logger>()
+        val settingsStore =
+            CoderSettingsStore(serviceLocator.getService<PluginSettingsStore>(), Environment(), delegate)
+        val logger = CoderLogger(delegate, ui, cs, i18n, settingsStore.readOnly())
         return CoderRemoteProvider(
             CoderToolboxContext(
                 ui,
@@ -41,12 +45,12 @@ class CoderToolboxExtension : RemoteDevExtension {
                 serviceLocator.getService<ClientHelper>(),
                 serviceLocator.getService<LocalDesktopManager>(),
                 cs,
-                serviceLocator.getService<Logger>(),
+                logger,
                 i18n,
-                CoderSettingsStore(serviceLocator.getService<PluginSettingsStore>(), Environment(), logger),
+                settingsStore,
                 CoderSecretsStore(serviceLocator.getService<PluginSecretStore>()),
                 serviceLocator.getService<ToolboxProxySettings>(),
             )
         )
     }
-}
+}

src/main/kotlin/com/coder/toolbox/cli/SshCommandProcessHandle.kt

@@ -26,11 +26,10 @@ class SshCommandProcessHandle(private val ctx: CoderToolboxContext) {
         while (stack.isNotEmpty()) {
             val processHandle = stack.removeLast()
             val cmdLine = processHandle.info().commandLine().getOrNull()
-            ctx.logger.debug(sessionId, "SSH command PID: ${processHandle.pid()} Command: $cmdLine")
             if (cmdLine != null && cmdLine.isSshCommandFor(ws, agent)) {
                 ctx.logger.debug(
                     sessionId,
-                    "SSH command with PID: ${processHandle.pid()} and Command: $cmdLine matches ${ws.name}.${agent.name}"
+                    "SSH command with PID: ${processHandle.pid()} matches ${ws.name}.${agent.name}"
                 )
                 return processHandle.pid()
             } else {