Merge lp:~le-chi-thu/lava-scheduler/loop-health-check-job into lp:lava-scheduler

[Le Chi Thu]

Added looping state for health check job. When in this state, the health check job will run forever. Admin should then put the device to maintaining mode to terminate the looping.

Scheduler device page is updated with a button to put the device into looping mode.

[Michael Hudson-Doyle (mwhudson)]

Nice!

I'm not sure having the button on the page is really necessary -- I would have been happy to use the admin interface for this, I think -- but since it's done, it's nice to have indeed.

review: Approve

=== modified file 'lava_scheduler_app/models.py'
--- lava_scheduler_app/models.py	2012-04-20 03:22:44 +0000
+++ lava_scheduler_app/models.py	2012-05-23 13:56:35 +0000
@@ -74,11 +74,12 @@
     )
 
     # A device health shows a device is ready to test or not
-    HEALTH_UNKNOWN, HEALTH_PASS, HEALTH_FAIL = range(3)
+    HEALTH_UNKNOWN, HEALTH_PASS, HEALTH_FAIL, HEALTH_LOOPING = range(4)
     HEALTH_CHOICES = (
         (HEALTH_UNKNOWN, 'Unknown'),
         (HEALTH_PASS, 'Pass'),
         (HEALTH_FAIL, 'Fail'),
+        (HEALTH_LOOPING, 'Looping'),
     )
 
     hostname = models.CharField(
@@ -147,6 +148,8 @@
             created_by=user, device=self, old_state=self.status,
             new_state=new_status, message=reason, job=None).save()
         self.status = new_status
+        if self.health_status == Device.HEALTH_LOOPING:
+            self.health_status = Device.HEALTH_UNKNOWN
         self.save()
 
     def put_into_online_mode(self, user, reason):
@@ -160,6 +163,17 @@
         self.health_status = Device.HEALTH_UNKNOWN
         self.save()
 
+    def put_into_looping_mode(self, user):
+        if self.status not in [Device.OFFLINE, Device.OFFLINING]:
+            return
+        new_status = self.IDLE
+        DeviceStateTransition.objects.create(
+            created_by=user, device=self, old_state=self.status,
+            new_state=new_status, message="Looping mode", job=None).save()
+        self.status = new_status
+        self.health_status = Device.HEALTH_LOOPING
+        self.save()
+
     #@classmethod
     #def find_devices_by_type(cls, device_type):
     #    return device_type.device_set.all()

=== modified file 'lava_scheduler_app/templates/lava_scheduler_app/_device_base.html'
--- lava_scheduler_app/templates/lava_scheduler_app/_device_base.html	2012-03-09 02:49:54 +0000
+++ lava_scheduler_app/templates/lava_scheduler_app/_device_base.html	2012-05-23 13:56:35 +0000
@@ -54,6 +54,11 @@
   {% csrf_token %}
   <input name="reason" style="width: 100%"/>
 </form>
+<form style="display:inline " method="POST"
+      action="{% url lava.scheduler.device.looping device.pk %}">
+  {% csrf_token %}
+  <button id="looping-button">Put into looping mode</button>
+</form>
 {% endif %}
 
 <div id="columns">

=== modified file 'lava_scheduler_app/urls.py'
--- lava_scheduler_app/urls.py	2012-03-02 04:35:51 +0000
+++ lava_scheduler_app/urls.py	2012-05-23 13:56:35 +0000
@@ -3,33 +3,36 @@
 
 urlpatterns = patterns(
     'lava_scheduler_app.views',
-    url(r'^$', 
+    url(r'^$',
         'index',
         name='lava.scheduler'),
-    url(r'^active_jobs_json$', 
+    url(r'^active_jobs_json$',
         'index_active_jobs_json',
         name='lava.scheduler.active_jobs_json'),
-    url(r'^devices_json$', 
+    url(r'^devices_json$',
         'index_devices_json',
         name='lava.scheduler.index_devices_json'),
-    url(r'^alljobs$', 
-        'job_list', 
+    url(r'^alljobs$',
+        'job_list',
         name='lava.scheduler.job.list'),
     url(r'^alljobs_json$',
         'alljobs_json',
         name='lava.scheduler.job.list_json'),
-    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)$', 
-        'device_detail', 
+    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)$',
+        'device_detail',
         name='lava.scheduler.device.detail'),
-    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/recent_jobs_json$', 
-        'recent_jobs_json', 
+    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/recent_jobs_json$',
+        'recent_jobs_json',
         name='lava.scheduler.device.recent_jobs_json'),
-    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/transition_json$', 
-        'transition_json', 
+    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/transition_json$',
+        'transition_json',
         name='lava.scheduler.device.transition_json'),
-    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/maintenance$', 
-        'device_maintenance_mode', 
+    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/maintenance$',
+        'device_maintenance_mode',
         name='lava.scheduler.device.maintenance'),
+    url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/looping$',
+        'device_looping_mode',
+        name='lava.scheduler.device.looping'),
     url(r'^device/(?P<pk>[-_a-zA-Z0-9]+)/online$',
         'device_online',
         name='lava.scheduler.device.online'),
@@ -64,7 +67,7 @@
         'job_cancel',
         name='lava.scheduler.job.cancel'),
     url(r'^job/(?P<pk>[0-9]+)/json$',
-        'job_json', 
+        'job_json',
         name='lava.scheduler.job.json'),
     url(r'^job/(?P<pk>[0-9]+)/output$',
         'job_output',

=== modified file 'lava_scheduler_app/views.py'
--- lava_scheduler_app/views.py	2012-03-19 02:01:59 +0000
+++ lava_scheduler_app/views.py	2012-05-23 13:56:35 +0000
@@ -573,3 +573,14 @@
     else:
         return HttpResponseForbidden(
             "you cannot administer this device", content_type="text/plain")
+
+@post_only
+def device_looping_mode(request, pk):
+    device = Device.objects.get(pk=pk)
+    if device.can_admin(request.user):
+        device.put_into_looping_mode(request.user)
+        return redirect(device)
+    else:
+        return HttpResponseForbidden(
+            "you cannot administer this device", content_type="text/plain")
+

=== modified file 'lava_scheduler_daemon/dbjobsource.py'
--- lava_scheduler_daemon/dbjobsource.py	2012-04-05 04:34:13 +0000
+++ lava_scheduler_daemon/dbjobsource.py	2012-05-23 13:56:35 +0000
@@ -152,6 +152,8 @@
                 run_health_check = False
             elif device.health_status == Device.HEALTH_UNKNOWN:
                 run_health_check = True
+            elif device.health_status == Device.HEALTH_LOOPING:
+                run_health_check = True
             elif not device.last_health_report_job:
                 run_health_check = True
             else:
@@ -238,11 +240,12 @@
 
         if job.health_check:
             device.last_health_report_job = job
-            if job.status == TestJob.INCOMPLETE:
-                device.health_status = Device.HEALTH_FAIL
-                device.put_into_maintenance_mode(None, "Health Check Job Failed")
-            elif job.status == TestJob.COMPLETE:
-                device.health_status = Device.HEALTH_PASS
+            if device.health_status != Device.HEALTH_LOOPING:
+                if job.status == TestJob.INCOMPLETE:
+                    device.health_status = Device.HEALTH_FAIL
+                    device.put_into_maintenance_mode(None, "Health Check Job Failed")
+                elif job.status == TestJob.COMPLETE:
+                    device.health_status = Device.HEALTH_PASS
 
         job.end_time = datetime.datetime.utcnow()
         token = job.submit_token