Merge lp:~josharenson/qa-dashboard/mir_performance into lp:qa-dashboard

[Josh Arenson]

Adds mir_performance application plugin

Adds mir_performance application plugin

[lp:~josharenson/qa-dashboard/mir_performance]

765. By Josh Arenson on 2014-06-23

Removed benchmarks.js

766. By Josh Arenson on 2014-06-23

Removed stray views.py

767. By Josh Arenson on 2014-06-23

Removed pyc files

768. By Josh Arenson on 2014-06-24

Fixed a forgotten import statement

769. By Josh Arenson on 2014-06-25

Added __get_master_job method as the one in the jenkinsapi doesn't work quite right

[lp:~josharenson/qa-dashboard/mir_performance]

770. By Josh Arenson on 2014-06-30

 * Added correct copyright notices
 * Disabled debug mode

771. By Josh Arenson on 2014-06-30

Fix style violations

[PS Jenkins bot (ps-jenkins)]

FAILED: Continuous integration, rev:767
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/329/
Executed test runs:

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/329/rebuild

review: Needs Fixing (continuous-integration)

[PS Jenkins bot (ps-jenkins)]

FAILED: Continuous integration, rev:771
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/332/
Executed test runs:

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/332/rebuild

review: Needs Fixing (continuous-integration)

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:771
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/336/
Executed test runs:

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/336/rebuild

review: Approve (continuous-integration)

[Chris Johnston (cjohnston)]

If you add the following line to your local_settings.py you should be able to use the local .js files and remove the inline JS..

STATIC_URL = '/static/'

[Josh Arenson (josharenson)]

> If you add the following line to your local_settings.py you should be able to
> use the local .js files and remove the inline JS..
>
> STATIC_URL = '/static/'

Is there any way you can make an exception? I am unable to test the site when I do this as it introduces external dependencies that do not work in my test environment.

[lp:~josharenson/qa-dashboard/mir_performance]

772. By Josh Arenson on 2014-08-14

Merged upstream

773. By Josh Arenson on 2014-08-20

Make the scale of the graphs such that they show less detail. This makes it easier to see sharp trends.

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:773
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/358/
Executed test runs:

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/358/rebuild

review: Approve (continuous-integration)

[lp:~josharenson/qa-dashboard/mir_performance]

774. By Josh Arenson on 2014-10-21

Fix homepage link for Mir Performance

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:774
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/375/
Executed test runs:

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dashboard-ci/375/rebuild

review: Approve (continuous-integration)

Unmerged revisions

774. By Josh Arenson on 2014-10-21

Fix homepage link for Mir Performance

773. By Josh Arenson on 2014-08-20

Make the scale of the graphs such that they show less detail. This makes it easier to see sharp trends.

772. By Josh Arenson on 2014-08-14

Merged upstream

771. By Josh Arenson on 2014-06-30

Fix style violations

770. By Josh Arenson on 2014-06-30

 * Added correct copyright notices
 * Disabled debug mode

769. By Josh Arenson on 2014-06-25

Added __get_master_job method as the one in the jenkinsapi doesn't work quite right

768. By Josh Arenson on 2014-06-24

Fixed a forgotten import statement

767. By Josh Arenson on 2014-06-23

Removed pyc files

766. By Josh Arenson on 2014-06-23

Removed stray views.py

765. By Josh Arenson on 2014-06-23

Removed benchmarks.js

=== added directory 'mir_performance'
=== added file 'mir_performance/__init__.py'
=== added file 'mir_performance/dashboard.py'
--- mir_performance/dashboard.py	1970-01-01 00:00:00 +0000
+++ mir_performance/dashboard.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,48 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from django.core.management import call_command
+from django.conf.urls import include, patterns, url
+from django.core.urlresolvers import reverse_lazy
+
+from common.plugin_helper import (
+    Extension,
+)
+
+
+class MirPerformanceExtension(Extension):
+    URL_PATTERNS = patterns('',
+            url(r'^mir_performance/', include('mir_performance.urls')),
+    )
+
+    LINKS = [
+        {
+            'name'  : 'mir_performance',
+            'url'   : reverse_lazy('glmark2'),
+            'label' : 'Mir Performance',
+            'index' : 500,
+        },
+    ]
+
+    def pull_data(self):
+        call_command('jenkins_pull_mir_performance', '--scrape')
+
+    @classmethod
+    def name(cls):
+        return cls.__name__
+
+
+extension = MirPerformanceExtension()
+extension.pull_data()

=== added directory 'mir_performance/management'
=== added file 'mir_performance/management/__init__.py'
=== added directory 'mir_performance/management/commands'
=== added file 'mir_performance/management/commands/README'
--- mir_performance/management/commands/README	1970-01-01 00:00:00 +0000
+++ mir_performance/management/commands/README	2014-10-21 13:09:04 +0000
@@ -0,0 +1,15 @@
+The flow of this is as follows:
+
+// --scrape example given, as --last-successful is a subset of that
+
+1. Connect to jenkins <build_host>
+
+2. Determine the last successful build number
+
+3. Get metadata for all artifacts from last {n} builds (default is 20)
+
+4. If the artifact filename indicated that it is a benchmark, as determined by <benchmark_artifact_suffix>, the artifact is saved for processing.
+
+5. PerformanceBenchmarkFactory returns an object that is capable of processing that specific artifact's data. Extending this class is how new benchmarks can be added.
+
+6. Whatever the factory returns is parsed, and if the result is unique, a Benchmark object (Django model) is saved into the database!

=== added file 'mir_performance/management/commands/_PerformanceBenchmarkFactory.py'
--- mir_performance/management/commands/_PerformanceBenchmarkFactory.py	1970-01-01 00:00:00 +0000
+++ mir_performance/management/commands/_PerformanceBenchmarkFactory.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,94 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import json
+import re
+
+from mir_performance.models import (
+    GLMARK2,
+    Benchmark
+)
+
+"""
+    A Boilerplate, polyymorphic factory class to create benchmark objects
+    based on the benchmark type
+"""
+class PerformanceBenchmarkFactory:
+    factories = {}
+    def addFactory(id, performanceBenchmarkFactory):
+        PerformanceBenchmarkFactory.factories.put[id] = performanceBenchmarkFactory
+    addFactory = staticmethod(addFactory)
+
+    def createPerformanceBenchmark(job_name, build_num, artifact, parent_benchmark_name):
+        pbf = PerformanceBenchmarkFactory
+        if not pbf.factories.has_key(parent_benchmark_name):
+            pbf.factories[parent_benchmark_name] = eval(parent_benchmark_name + '.Factory()')
+        return pbf.factories[parent_benchmark_name].create(job_name, build_num, artifact, parent_benchmark_name)
+    createPerformanceBenchmark = staticmethod(createPerformanceBenchmark)
+
+class PerformanceBenchmark(object):
+    def __init__(self, job_name, build_num, artifact, parent_benchmark_name):
+        self.job_name = job_name
+        self.build_num = build_num
+        self.artifact = artifact
+        self.parent_benchmark_name = parent_benchmark_name
+
+    def insert_artifact_data(self):
+        raise NotImplementedError
+
+class glmark2_fullscreen_default(PerformanceBenchmark):
+
+    def insert_artifact_data(self):
+        benchmarks = self.__get_benchmark_objects(self.artifact)
+        for benchmark in benchmarks:
+            # Do we already have a benchmark with this name && build number?
+            done_this_build = Benchmark.objects.filter(jenkins_build=self.build_num).filter(name=benchmark.name)
+            if not done_this_build:
+                benchmark.save()
+            else:
+                continue
+                #print "Already had GLMark2 data for build %d" % self.build_num
+
+    def __get_benchmark_objects(self, artifact):
+        result = []
+        re_benchmark_name = re.compile("(?P<name>^.*?)\:\s+FPS\:\s+(?P<score>.*)\sFrameTime")
+        re_overall_score = "(?P<name>glmark2\s+Score)\:\s+(?P<score>\d+)"
+
+        stream = artifact.get_data()
+        for line in stream.splitlines():
+            b = None
+            match = re.search(re_benchmark_name, line)
+            if match:
+                b = Benchmark() # our django model
+                b.parent = GLMARK2
+                b.name = match.group('name')
+                b.score = match.group('score')
+                b.jenkins_build = int(self.build_num)
+            else:
+                match = re.search(re_overall_score, line)
+                if match:
+                    b = Benchmark()
+                    b.parent = GLMARK2
+                    b.name = match.group('name')
+                    b.score = match.group('score')
+                    b.jenkins_build = int(self.build_num)
+            if b is not None:
+                result.append(b)
+        return result
+
+    class Factory:
+        def create(self, job_name, build_num, artifact, parent_benchmark_name):
+            return glmark2_fullscreen_default(job_name, build_num, artifact, parent_benchmark_name)
+

=== added file 'mir_performance/management/commands/__init__.py'
=== added file 'mir_performance/management/commands/jenkins_pull_mir_performance.py'
--- mir_performance/management/commands/jenkins_pull_mir_performance.py	1970-01-01 00:00:00 +0000
+++ mir_performance/management/commands/jenkins_pull_mir_performance.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,122 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import jenkinsapi
+from jenkinsapi.build import Build
+from jenkinsapi.jenkins import Jenkins
+from jenkinsapi.jenkins import Job
+from optparse import make_option
+import traceback
+
+from django.core.management.base import BaseCommand, CommandError
+
+import _PerformanceBenchmarkFactory as Pbf
+from mir_performance.models import Benchmark
+
+DBG = False
+
+build_host = "https://jenkins.qa.ubuntu.com"
+
+benchmark_artifact_suffix = ".results"
+job_names = ['mir-mediumtests-runner-mako']
+
+
+class Command(BaseCommand):
+    help = "Pull mir performance information from jenkins into the database"
+    option_list = BaseCommand.option_list + (
+        make_option('--scrape',
+            action='store_true',
+            dest='scrape',
+            default=False,
+            help='Scrape the results from the last 20 builds.'
+            ),
+        make_option('--last-successful',
+            action='store_true',
+            dest='last_successful',
+            default=False,
+            help='Get results from the last successful build'),
+        )
+
+    def handle(self, *args, **options):
+        try:
+            self.J = Jenkins(build_host)
+            if options['scrape']:
+                self.scrape()
+            if options['last_successful']:
+                self.get_last_successful()
+        except:
+            if DBG:
+                traceback.print_exc()
+
+    def get_last_successful(self):
+        self.scrape(1)
+
+    def scrape(self, build_nums_back=20):
+        for job_name in job_names:
+            most_recent = self.__last_successful_job_num(job_name)
+
+            # Get a single build's artifacts
+            for build_num in range(most_recent - build_nums_back,
+                                   most_recent):
+                artifacts = self.__get_benchmark_artifacts(build_num, job_name)
+                # Create appropriate object based on artifact filename
+                for artifact in artifacts:
+                    self.__insert_artifact(job_name, build_num, artifact)
+
+
+    """
+        Returns: list of Artifact objects whose filenames
+        end with <benchmark_artifact_suffix>
+    """
+    def __get_benchmark_artifacts(self, build_num, job_name):
+        results = []
+        job = self.J.get_job(job_name)
+        build = job.get_build(build_num)
+        artifacts = build.get_artifacts()
+        for artifact in artifacts:
+            if artifact.filename.endswith(benchmark_artifact_suffix):
+                print "Found benchmark artifact %s" % artifact.filename
+                results.append(artifact)
+        return results
+
+
+    def __insert_artifact(self, job_name, build_num, artifact):
+        try:
+            # By convention, the parent benchmark name is the part of the
+            # filename before the suffix
+            fname = artifact.filename
+            parent_benchmark = fname[:fname.index(benchmark_artifact_suffix)]
+            benchmark = Pbf.PerformanceBenchmarkFactory.createPerformanceBenchmark(
+                job_name, build_num, artifact, parent_benchmark)
+            benchmark.insert_artifact_data()
+        except:
+            raise
+
+    def __get_master_job(self, child_job_name):
+        print "entry child: %s" % child_job_name
+        if child_job_name in self.J:
+            build_num = self.J[child_job_name].get_last_good_buildnumber()
+            parent = self.J[child_job_name].get_build(build_num).get_upstream_job()
+            if parent is not None:
+                return self.__get_master_job(str(parent))
+            else:
+                return child_job_name
+
+
+    def __last_successful_job_num(self, job_name="mir-mediumtests-runner-mako"):
+        last_good = self.J[job_name].get_last_good_buildnumber()
+        print "Branch %" % self.J[job_name].get_build(last_good).get_revision_branch()
+        return last_good
+

=== added file 'mir_performance/models.py'
--- mir_performance/models.py	1970-01-01 00:00:00 +0000
+++ mir_performance/models.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,33 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+
+from django.db import models
+
+from common.models import DashboardBaseModel
+
+UNKNOWN = 0
+GLMARK2 = 1
+PARENT_BENCHMARK_CHOICES = (
+    (UNKNOWN, 'Unknown'),
+    (GLMARK2, 'GLMark2'),
+)
+
+class Benchmark(DashboardBaseModel):
+    parent = models.SmallIntegerField(choices=PARENT_BENCHMARK_CHOICES)
+    name = models.CharField(max_length=1024)
+    score = models.IntegerField()
+    jenkins_build = models.IntegerField()
+

=== added directory 'mir_performance/static'
=== added directory 'mir_performance/static/mir_performance'
=== added directory 'mir_performance/static/mir_performance/js'
=== added directory 'mir_performance/templates'
=== added directory 'mir_performance/templates/mir_performance'
=== added file 'mir_performance/templates/mir_performance/mako_results.html'
--- mir_performance/templates/mir_performance/mako_results.html	1970-01-01 00:00:00 +0000
+++ mir_performance/templates/mir_performance/mako_results.html	2014-10-21 13:09:04 +0000
@@ -0,0 +1,82 @@
+{% extends "mir_performance/mir_performance_layout.html" %}
+{% load staticfiles %}
+
+{% block page_name %}Mir performance{% endblock %}
+
+{% block content %}
+<style>
+    {% for name, atts in benchmark_listi.items  %}
+    {# since the benchmark name is garbage, we use a simple index #}
+        #chart-{{ forloop.counter }} svg {
+            height: 600px;
+        }
+    {% endfor %}
+</style>
+
+<script>
+
+	function draw_benchmark_score_over_time_chart(name, atts, chart_id){
+		nv.addGraph(function() {
+            var chart = nv.models.lineChart()
+			    .margin({left: 100})
+				.showYAxis(true)
+				.showXAxis(true)
+		    ;
+
+		    chart.xAxis
+                .axisLabel("Build Number")
+
+		    chart.yAxis
+			    .axisLabel('Score')
+                .scale().domain([0,100])
+
+		    var myData = get_chart_data(atts);
+
+		    d3.select('#chart-' + chart_id + ' svg') 
+                .datum(myData)
+			    .call(chart)
+            ;
+
+            d3.select('#chart-' + chart_id + ' div')
+                .text(name)
+            ;
+
+
+		  nv.utils.windowResize(function() { chart.update() });
+		  return chart;
+		});
+	}
+
+	function get_chart_data(atts) {
+        snb = [];
+        for(var i = 0; i < atts.length; i++){
+            snb.push({x: atts[i].build_num, y: atts[i].score})
+        }
+
+	  //Line chart data should be sent as an array of series objects.
+	  return [
+		{
+		  values: snb,
+		  key: 'Score Trend',
+		  color: '#dd4814'
+		},
+	  ];
+	}
+
+    {% autoescape off %}
+    {% for name, atts in benchmark_list.items %}
+    draw_benchmark_score_over_time_chart("{{ name|escape }}", {{ atts }}, {{forloop.counter}})
+    {% endfor %}
+    {% endautoescape %}
+</script>
+
+<div>
+    {% for name, atts in benchmark_list.items %}
+    <div id="chart-{{ forloop.counter }}">
+        <div style='text-align:center;padding-top:10px'></div>
+        <svg></svg>
+    </div>
+    {% endfor %}
+</div>
+
+{% endblock %}

=== added file 'mir_performance/templates/mir_performance/mir_performance_layout.html'
--- mir_performance/templates/mir_performance/mir_performance_layout.html	1970-01-01 00:00:00 +0000
+++ mir_performance/templates/mir_performance/mir_performance_layout.html	2014-10-21 13:09:04 +0000
@@ -0,0 +1,13 @@
+{% extends "layout.html" %}
+
+{% load staticfiles %}
+
+{% block extra_headers %}
+    <link href='http://6df403e3d98e2ac67ac2-180150c581869d2c4c18db9c9e3179c4.r40.cf1.rackcdn.com/nv.d3.css' rel='stylesheet' type='text/css' /    >
+    <script type='text/javascript' src='//cdnjs.cloudflare.com/ajax/libs/d3/3.1.6/d3.min.js'></script>
+    <script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/nvd3/1.0.0-beta/nv.d3.min.js"></script>
+{% endblock %}
+
+{% block sub_nav_links %}
+<li {% ifequal url.1 'glmark2' %}class="active"{% endifequal %}><a class="sub-nav-item" href='{% url "glmark2" %}'>GLMark2</a></li>
+{% endblock %}

=== added file 'mir_performance/tests.py'
--- mir_performance/tests.py	1970-01-01 00:00:00 +0000
+++ mir_performance/tests.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,16 @@
+"""
+This file demonstrates writing tests using the unittest module. These will pass
+when you run "manage.py test".
+
+Replace this with more appropriate tests for your application.
+"""
+
+from django.test import TestCase
+
+
+class SimpleTest(TestCase):
+    def test_basic_addition(self):
+        """
+        Tests that 1 + 1 always equals 2.
+        """
+        self.assertEqual(1 + 1, 2)

=== added file 'mir_performance/urls.py'
--- mir_performance/urls.py	1970-01-01 00:00:00 +0000
+++ mir_performance/urls.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,21 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+from django.conf.urls import patterns, url
+
+urlpatterns = patterns(
+    'mir_performance.views',
+    url('^glmark2/$', 'glmark2', name='glmark2'),
+)

=== added file 'mir_performance/views.py'
--- mir_performance/views.py	1970-01-01 00:00:00 +0000
+++ mir_performance/views.py	2014-10-21 13:09:04 +0000
@@ -0,0 +1,70 @@
+# QA Dashboard
+# Copyright 2014 Canonical Ltd.
+
+# This program is free software: you can redistribute it and/or modify it
+# under the terms of the GNU Affero General Public License version 3, as
+# published by the Free Software Foundation.
+
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranties of
+# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
+# PURPOSE.  See the GNU Affero General Public License for more details.
+
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import json
+import os
+
+from django.http import HttpResponse
+from django.template import RequestContext
+from django.shortcuts import render_to_response
+
+from common.views import index
+
+from mir_performance.models import (
+    GLMARK2,
+    Benchmark
+)
+
+def sample_data(request):
+    return
+
+def glmark2(request):
+    # Rather than just limit the results returned by the query, we have to
+    # do some work.
+    # TODO rearchitect this to be better
+    data = {}
+    data = organize_benchmarks(GLMARK2)
+
+    page = {'benchmark_list' : data}
+    return render_to_response('mir_performance/mako_results.html',page,RequestContext(request))
+
+
+"""
+TRICKY: This method returns a list of json objects, where each object is a list of
+size <data_points>. Each inner object is a group of the same children benchmarks.
+This allows a template for each parent benchmark to display data from all children benchmarks
+looking back over <data_points> number of builds.
+"""
+def organize_benchmarks(benchmark_parent_id, data_points=20):
+    # First, get the most recent.
+    children_list = {}
+    top_res = Benchmark.objects.filter(parent=benchmark_parent_id).order_by('-jenkins_build')[0]
+    newest = top_res.jenkins_build
+    oldest = newest - data_points
+
+    for build_num in range(oldest,newest):
+        children = Benchmark.objects.filter(jenkins_build=build_num).filter(parent=benchmark_parent_id)
+
+        for child in children:
+            if not child.name in children_list:
+                children_list[child.name] = []
+            t = {}
+            t['score'] = child.score
+            t['build_num'] = child.jenkins_build
+            children_list[child.name].append(t)
+    return children_list
+
+
+