Mitigate concurrency issues

.github/workflows/test-using-pytest.yml

@@ -52,7 +52,8 @@ jobs:
           chmod +x /opt/firecracker/firecracker
           chmod +x /opt/firecracker/jailer
 
-          find /opt
+          # this produces a 33 MB log
+          #  find /opt
 
       - name: "Build custom runtimes"
         run: |
@@ -71,6 +72,11 @@ jobs:
         run: |
           sudo python3 -m pip install hatch hatch-vcs coverage
           sudo hatch run testing:cov
+      - name: Output modules used and their version
+        if:  always()
+        run: |
+           sudo hatch -e testing run pip freeze
+
 
       - name: Upload coverage reports to Codecov
         uses: codecov/[email protected]

pyproject.toml

@@ -87,6 +87,7 @@ check = "aleph-vm controller run {args:--help}"
 type = "virtual"
 system-packages = true
 dependencies = [
+  "eth_typing==4.3.1", # Temp fix for bug in CI with 5.0.0
   "pytest==8.2.1",
   "pytest-cov==5.0.0",
   "pytest-mock==3.14.0",

src/aleph/vm/controllers/qemu/instance.py

@@ -166,7 +166,7 @@ async def configure(self):
         """Configure the VM by saving controller service configuration"""
 
         logger.debug(f"Making  Qemu configuration: {self} ")
-        monitor_socket_path = settings.EXECUTION_ROOT / (str(self.vm_id) + "-monitor.socket")
+        monitor_socket_path = settings.EXECUTION_ROOT / (str(self.vm_hash) + "-monitor.socket")
 
         cloud_init_drive = await self._create_cloud_init_drive()
 
@@ -218,7 +218,7 @@ def save_controller_configuration(self):
 
     @property
     def qmp_socket_path(self) -> Path:
-        return settings.EXECUTION_ROOT / f"{self.vm_id}-qmp.socket"
+        return settings.EXECUTION_ROOT / f"{self.vm_hash}-qmp.socket"
 
     async def start(self):
         # Start via systemd not here

src/aleph/vm/models.py

@@ -96,6 +96,10 @@ def is_running(self) -> bool:
             else bool(self.times.starting_at and not self.times.stopping_at)
         )
 
+    @property
+    def is_allocated(self) -> bool:
+        return bool(self.times.starting_at and not self.times.stopping_at)
+
     @property
     def is_stopping(self) -> bool:
         return bool(self.times.stopping_at and not self.times.stopped_at)

src/aleph/vm/pool.py

@@ -161,7 +161,9 @@ def get_unique_vm_id(self) -> int:
             #
             # We therefore recycle vm_id values from executions that are not running
             # anymore.
-            currently_used_vm_ids = {execution.vm_id for execution in self.executions.values() if execution.is_running}
+            currently_used_vm_ids = {
+                execution.vm_id for execution in self.executions.values() if execution.is_allocated
+            }
             for i in range(settings.START_ID_INDEX, 255**2):
                 if i not in currently_used_vm_ids:
                     return i