merge: Add tests for interference and failures in Capelin

This pull request updates the Capelin experiments to test for interference and failure scenarios. 
This allows us to track regressions in these subsystems more easily.

* Clean up Bitbrains trace reader to enable re-use 
* Keep trace order after sampling
* Update Bitbrains trace tests
* Add support for reporting interfered work
* Add support for SimHost failure
* Add tests for interference and failures

1 files changed, 21 insertions, 5 deletions

diff --git a/opendc-compute/opendc-compute-simulator/src/main/kotlin/org/opendc/compute/simulator/SimHost.kt b/opendc-compute/opendc-compute-simulator/src/main/kotlin/org/opendc/compute/simulator/SimHost.kt
index 5ea577f3..be771f6d 100644
--- a/opendc-compute/opendc-compute-simulator/src/main/kotlin/org/opendc/compute/simulator/SimHost.kt
+++ b/opendc-compute/opendc-compute-simulator/src/main/kotlin/org/opendc/compute/simulator/SimHost.kt
@@ -46,6 +46,7 @@ import org.opendc.simulator.resources.SimResourceInterpreter
 import java.util.*
 import kotlin.coroutines.CoroutineContext
 import kotlin.coroutines.resume
+import kotlin.coroutines.resumeWithException
 
 /**
  * A [Host] that is simulates virtual machines on a physical machine using [SimHypervisor].
@@ -315,10 +316,16 @@ public class SimHost(
 
     override suspend fun fail() {
         _state = HostState.DOWN
+        for (guest in guests.values) {
+            guest.fail()
+        }
     }
 
     override suspend fun recover() {
         _state = HostState.UP
+        for (guest in guests.values) {
+            guest.start()
+        }
     }
 
     /**
@@ -329,7 +336,7 @@ public class SimHost(
 
         suspend fun start() {
             when (state) {
-                ServerState.TERMINATED -> {
+                ServerState.TERMINATED, ServerState.ERROR -> {
                     logger.info { "User requested to start server ${server.uid}" }
                     launch()
                 }
@@ -356,9 +363,15 @@ public class SimHost(
 
         suspend fun terminate() {
             stop()
+            machine.close()
             state = ServerState.DELETED
         }
 
+        suspend fun fail() {
+            stop()
+            state = ServerState.ERROR
+        }
+
         private var job: Job? = null
 
         private suspend fun launch() = suspendCancellableCoroutine<Unit> { cont ->
@@ -366,16 +379,19 @@ public class SimHost(
             val workload = mapper.createWorkload(server)
 
             job = scope.launch {
-                delay(1) // TODO Introduce boot time
-                init()
-                cont.resume(Unit)
+                try {
+                    delay(1) // TODO Introduce boot time
+                    init()
+                    cont.resume(Unit)
+                } catch (e: Throwable) {
+                    cont.resumeWithException(e)
+                }
                 try {
                     machine.run(workload, mapOf("driver" to this@SimHost, "server" to server))
                     exit(null)
                 } catch (cause: Throwable) {
                     exit(cause)
                 } finally {
-                    machine.close()
                     job = null
                 }
             }