Add metric for image build result

Add a metric that shows on a per-image basis whether an image build was
successful or not.

Change-Id: I8e97017dd3f91cebef3791168371b29899b83389

doc/source/operation.rst

@@ -294,6 +294,16 @@ Nodepool builder
    Number of image uploads to a specific provider in the cloud plus the time in
    seconds spent to upload the image.
 
+.. zuul:stat:: nodepool.builder.dib_image_build.<diskimage_name>.<ext>.rc
+   :type: gauge
+
+   Return code of the DIB.
+
+.. zuul:stat:: nodepool.builder.dib_image_build.<diskimage_name>.<ext>.duration
+   :type: timer
+
+   Time the DIB run took.
+
 Nodepool launcher
 ~~~~~~~~~~~~~~~~~
 

nodepool/builder.py

@@ -749,6 +749,7 @@ class BuildWorker(BaseWorker):
         self.log.info('Running %s' % (cmd,))
         self.log.info('Logging to %s' % (log_fn,))
 
+        start_time = time.monotonic()
         try:
             p = subprocess.Popen(
                 shlex.split(cmd),
@@ -784,6 +785,8 @@ class BuildWorker(BaseWorker):
         if did_suspend:
             self.log.info("ZooKeeper available. Resuming")
 
+        build_time = time.monotonic() - start_time
+
         build_data = zk.ImageBuild()
         build_data.builder_id = self._builder_id
         build_data.builder = self._hostname
@@ -817,6 +820,15 @@ class BuildWorker(BaseWorker):
                                    (diskimage.name, filename, ext, size))
                     self._statsd.gauge(key, size)
 
+        if self._statsd:
+            # report result to statsd
+            for ext in img_types.split(','):
+                key_base = 'nodepool.dib_image_build.%s.%s' % (
+                    diskimage.name, ext)
+                self._statsd.gauge(key_base + '.rc', rc)
+                self._statsd.timing(key_base + '.duration',
+                                    int(build_time * 1000))
+
         return build_data
 
     def run(self):

nodepool/tests/test_builder.py

@@ -307,6 +307,10 @@ class TestNodePoolBuilder(tests.DBTestCase):
         self.waitForImage('fake-provider', 'fake-image')
         # Make sure our cleanup worker properly removes the first build.
         self.waitForBuildDeletion('fake-image', '0000000001')
+        self.assertReportedStat('nodepool.dib_image_build.fake-image.qcow2.rc',
+                                '127', 'g')
+        self.assertReportedStat('nodepool.dib_image_build.'
+                                'fake-image.qcow2.duration', None, 'ms')
 
     def test_diskimage_build_only(self):
         configfile = self.setup_config('node_diskimage_only.yaml')
@@ -317,6 +321,10 @@ class TestNodePoolBuilder(tests.DBTestCase):
 
         self.assertEqual(build_tar._formats, ['tar'])
         self.assertEqual(build_default._formats, ['qcow2'])
+        self.assertReportedStat('nodepool.dib_image_build.fake-image.tar.rc',
+                                '0', 'g')
+        self.assertReportedStat('nodepool.dib_image_build.'
+                                'fake-image.tar.duration', None, 'ms')
 
     def test_diskimage_build_formats(self):
         configfile = self.setup_config('node_diskimage_formats.yaml')

releasenotes/notes/build-result-metrics-deb2aaa329830f8a.yaml

@@ -0,0 +1,9 @@
+---
+features:
+  - |
+    Two new metrics are now reported after each run of the diskimage builder:
+    nodepool.builder.dib_image_build.<diskimage_name>.<ext>.rc will be set to
+    the last result code of the diskimage builder. This metric can be used to
+    set up alerting for failed disk image builds.
+    nodepool.builder.dib_image_build.<diskimage_name>.<ext>.duration will
+    receive the time it took to build the disk image.