# Copyright (c) 2018 Cisco and/or its affiliates.
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at:
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Generation of Continuous Performance Trending and Analysis.
"""
import multiprocessing
import os
import logging
import csv
import prettytable
import plotly.offline as ploff
import plotly.graph_objs as plgo
import plotly.exceptions as plerr
from collections import OrderedDict
from datetime import datetime
from utils import archive_input_data, execute_command, \
classify_anomalies, Worker
# Command to build the html format of the report
HTML_BUILDER = 'sphinx-build -v -c conf_cpta -a ' \
'-b html -E ' \
'-t html ' \
'-D version="{date}" ' \
'{working_dir} ' \
'{build_dir}/'
# .css file for the html format of the report
THEME_OVERRIDES = """/* override table width restrictions */
.wy-nav-content {
max-width: 1200px !important;
}
.rst-content blockquote {
margin-left: 0px;
line-height: 18px;
margin-bottom: 0px;
}
.wy-menu-vertical a {
display: inline-block;
line-height: 18px;
padding: 0 2em;
display: block;
position: relative;
font-size: 90%;
color: #d9d9d9
}
.wy-menu-vertical li.current a {
color: gray;
border-right: solid 1px #c9c9c9;
padding: 0 3em;
}
.wy-menu-vertical li.toctree-l2.current > a {
background: #c9c9c9;
padding: 0 3em;
}
.wy-menu-vertical li.toctree-l2.current li.toctree-l3 > a {
display: block;
background: #c9c9c9;
padding: 0 4em;
}
.wy-menu-vertical li.toctree-l3.current li.toctree-l4 > a {
display: block;
background: #bdbdbd;
padding: 0 5em;
}
.wy-menu-vertical li.on a, .wy-menu-vertical li.current > a {
color: #404040;
padding: 0 2em;
font-weight: bold;
position: relative;
background: #fcfcfc;
border: none;
border-top-width: medium;
border-bottom-width: medium;
border-top-style: none;
border-bottom-style: none;
border-top-color: currentcolor;
border-bottom-color: currentcolor;
padding-left: 2em -4px;
}
"""
COLORS = ["SkyBlue", "Olive", "Purple", "Coral", "Indigo", "Pink",
"Chocolate", "Brown", "Magenta", "Cyan", "Orange", "Black",
"Violet", "Blue", "Yellow", "BurlyWood", "CadetBlue", "Crimson",
"DarkBlue", "DarkCyan", "DarkGreen", "Green", "GoldenRod",
"LightGreen", "LightSeaGreen", "LightSkyBlue", "Maroon",
"MediumSeaGreen", "SeaGreen", "LightSlateGrey",
"SkyBlue", "Olive", "Purple", "Coral", "Indigo", "Pink",
"Chocolate", "Brown", "Magenta", "Cyan", "Orange", "Black",
"Violet", "Blue", "Yellow", "BurlyWood", "CadetBlue", "Crimson",
"DarkBlue", "DarkCyan", "DarkGreen", "Green", "GoldenRod",
"LightGreen", "LightSeaGreen", "LightSkyBlue", "Maroon",
"MediumSeaGreen", "SeaGreen", "LightSlateGrey"
]
def generate_cpta(spec, data):
"""Generate all formats and versions of the Continuous Performance Trending
and Analysis.
:param spec: Specification read from the specification file.
:param data: Full data set.
:type spec: Specification
:type data: InputData
"""
logging.info("Generating the Continuous Performance Trending and Analysis "
"...")
ret_code = _generate_all_charts(spec, data)
cmd = HTML_BUILDER.format(
date=datetime.utcnow().strftime('%Y-%m-%d %H:%M UTC'),
working_dir=spec.environment["paths"]["DIR[WORKING,SRC]"],
build_dir=spec.environment["paths"]["DIR[BUILD,HTML]"])
execute_command(cmd)
with open(spec.environment["paths"]["DIR[CSS_PATCH_FILE]"], "w") as \
css_file:
css_file.write(THEME_OVERRIDES)
with open(spec.environment["paths"]["DIR[CSS_PATCH_FILE2]"], "w") as \
css_file:
css_file.write(THEME_OVERRIDES)
archive_input_data(spec)
logging.info("Done.")
return ret_code
def _generate_trending_traces(in_data, job_name, build_info,
show_trend_line=True, name="", color=""):
"""Generate the trending traces:
- samples,
- outliers, regress, progress
- average of normal samples (trending line)
:param in_data: Full data set.
:param job_name: The name of job which generated the data.
:param build_info: Information about the builds.
:param show_trend_line: Show moving median (trending plot).
:param name: Name of the plot
:param color: Name of the color for the plot.
:type in_data: OrderedDict
:type job_name: str
:type build_info: dict
:type show_trend_line: bool
:type name: str
:type color: str
:returns: Generated traces (list) and the evaluated result.
:rtype: tuple(traces, result)
"""
data_x = list(in_data.keys())
data_y = list(in_data.values())
hover_text = list()
xaxis = list()
for idx in data_x:
date = build_info[job_name][str(idx)][0]
hover_str = ("date: {0}
"
"value: {1:,}
"
"{2}-ref: {3}
"
"csit-ref: mrr-{4}-build-{5}")
if "dpdk" in job_name:
hover_text.append(hover_str.format(
date,
int(in_data[idx].avg),
"dpdk",
build_info[job_name][str(idx)][1].
rsplit('~', 1)[0],
"weekly",
idx))
elif "vpp" in job_name:
hover_text.append(hover_str.format(
date,
int(in_data[idx].avg),
"vpp",
build_info[job_name][str(idx)][1].
rsplit('~', 1)[0],
"daily",
idx))
xaxis.append(datetime(int(date[0:4]), int(date[4:6]), int(date[6:8]),
int(date[9:11]), int(date[12:])))
data_pd = OrderedDict()
for key, value in zip(xaxis, data_y):
data_pd[key] = value
anomaly_classification, avgs = classify_anomalies(data_pd)
anomalies = OrderedDict()
anomalies_colors = list()
anomalies_avgs = list()
anomaly_color = {
"regression": 0.0,
"normal": 0.5,
"progression": 1.0
}
if anomaly_classification:
for idx, (key, value) in enumerate(data_pd.iteritems()):
if anomaly_classification[idx] in \
("outlier", "regression", "progression"):
anomalies[key] = value
anomalies_colors.append(
anomaly_color[anomaly_classification[idx]])
anomalies_avgs.append(avgs[idx])
anomalies_colors.extend([0.0, 0.5, 1.0])
# Create traces
trace_samples = plgo.Scatter(
x=xaxis,
y=[y.avg for y in data_y],
mode='markers',
line={
"width": 1
},
showlegend=True,
legendgroup=name,
name="{name}".format(name=name),
marker={
"size": 5,
"color": color,
"symbol": "circle",
},
text=hover_text,
hoverinfo="text"
)
traces = [trace_samples, ]
if show_trend_line:
trace_trend = plgo.Scatter(
x=xaxis,
y=avgs,
mode='lines',
line={
"shape": "linear",
"width": 1,
"color": color,
},
showlegend=False,
legendgroup=name,
name='{name}'.format(name=name),
text=["trend: {0:,}".format(int(avg)) for avg in avgs],
hoverinfo="text+name"
)
traces.append(trace_trend)
trace_anomalies = plgo.Scatter(
x=anomalies.keys(),
y=anomalies_avgs,
mode='markers',
hoverinfo="none",
showlegend=False,
legendgroup=name,
name="{name}-anomalies".format(name=name),
marker={
"size": 15,
"symbol": "circle-open",
"color": anomalies_colors,
"colorscale": [[0.00, "red"],
[0.33, "red"],
[0.33, "white"],
[0.66, "white"],
[0.66, "green"],
[1.00, "green"]],
"showscale": True,
"line": {
"width": 2
},
"colorbar": {
"y": 0.5,
"len": 0.8,
"title": "Circles Marking Data Classification",
"titleside": 'right',
"titlefont": {
"size": 14
},
"tickmode": 'array',
"tickvals": [0.167, 0.500, 0.833],
"ticktext": ["Regression", "Normal", "Progression"],
"ticks": "",
"ticklen": 0,
"tickangle": -90,
"thickness": 10
}
}
)
traces.append(trace_anomalies)
if anomaly_classification:
return traces, anomaly_classification[-1]
else:
return traces, None
def _generate_all_charts(spec, input_data):
"""Generate all charts specified in the specification file.
:param spec: Specification.
:param input_data: Full data set.
:type spec: Specification
:type input_data: InputData
"""
def _generate_chart(_, data_q, graph):
"""Generates the chart.
"""
logs = list()
logging.info(" Generating the chart '{0}' ...".
format(graph.get("title", "")))
logs.append(("INFO", " Generating the chart '{0}' ...".
format(graph.get("title", ""))))
job_name = graph["data"].keys()[0]
csv_tbl = list()
res = list()
# Transform the data
logs.append(("INFO", " Creating the data set for the {0} '{1}'.".
format(graph.get("type", ""), graph.get("title", ""))))
data = input_data.filter_data(graph, continue_on_error=True)
if data is None:
logging.error("No data.")
return
chart_data = dict()
for job, job_data in data.iteritems():
if job != job_name:
continue
for index, bld in job_data.items():
for test_name, test in bld.items():
if chart_data.get(test_name, None) is None:
chart_data[test_name] = OrderedDict()
try:
chart_data[test_name][int(index)] = \
test["result"]["receive-rate"]
except (KeyError, TypeError):
pass
# Add items to the csv table:
for tst_name, tst_data in chart_data.items():
tst_lst = list()
for bld in builds_dict[job_name]:
itm = tst_data.get(int(bld), '')
if not isinstance(itm, str):
itm = itm.avg
tst_lst.append(str(itm))
csv_tbl.append("{0},".format(tst_name) + ",".join(tst_lst) + '\n')
# Generate traces:
traces = list()
index = 0
for test_name, test_data in chart_data.items():
if not test_data:
logs.append(("WARNING", "No data for the test '{0}'".
format(test_name)))
continue
message = "index: {index}, test: {test}".format(
index=index, test=test_name)
test_name = test_name.split('.')[-1]
try:
trace, rslt = _generate_trending_traces(
test_data,
job_name=job_name,
build_info=build_info,
name='-'.join(test_name.split('-')[2:-1]),
color=COLORS[index])
except IndexError:
message = "Out of colors: {}".format(message)
logs.append(("ERROR", message))
logging.error(message)
index += 1
continue
traces.extend(trace)
res.append(rslt)
index += 1
if traces:
# Generate the chart:
graph["layout"]["xaxis"]["title"] = \
graph["layout"]["xaxis"]["title"].format(job=job_name)
name_file = "{0}-{1}{2}".format(spec.cpta["output-file"],
graph["output-file-name"],
spec.cpta["output-file-type"])
logs.append(("INFO", " Writing the file '{0}' ...".
format(name_file)))
plpl = plgo.Figure(data=traces, layout=graph["layout"])
try:
ploff.plot(plpl, show_link=False, auto_open=False,
filename=name_file)
except plerr.PlotlyEmptyDataError:
logs.append(("WARNING", "No data for the plot. Skipped."))
data_out = {
"job_name": job_name,
"csv_table": csv_tbl,
"results": res,
"logs": logs
}
data_q.put(data_out)
builds_dict = dict()
for job in spec.input["builds"].keys():
if builds_dict.get(job, None) is None:
builds_dict[job] = list()
for build in spec.input["builds"][job]:
status = build["status"]
if status != "failed" and status != "not found":
builds_dict[job].append(str(build["build"]))
# Create "build ID": "date" dict:
build_info = dict()
for job_name, job_data in builds_dict.items():
if build_info.get(job_name, None) is None:
build_info[job_name] = OrderedDict()
for build in job_data:
build_info[job_name][build] = (
input_data.metadata(job_name, build).get("generated", ""),
input_data.metadata(job_name, build).get("version", "")
)
work_queue = multiprocessing.JoinableQueue()
manager = multiprocessing.Manager()
data_queue = manager.Queue()
cpus = multiprocessing.cpu_count()
workers = list()
for cpu in range(cpus):
worker = Worker(work_queue,
data_queue,
_generate_chart)
worker.daemon = True
worker.start()
workers.append(worker)
os.system("taskset -p -c {0} {1} > /dev/null 2>&1".
format(cpu, worker.pid))
for chart in spec.cpta["plots"]:
work_queue.put((chart, ))
work_queue.join()
anomaly_classifications = list()
# Create the header:
csv_tables = dict()
for job_name in builds_dict.keys():
if csv_tables.get(job_name, None) is None:
csv_tables[job_name] = list()
header = "Build Number:," + ",".join(builds_dict[job_name]) + '\n'
csv_tables[job_name].append(header)
build_dates = [x[0] for x in build_info[job_name].values()]
header = "Build Date:," + ",".join(build_dates) + '\n'
csv_tables[job_name].append(header)
versions = [x[1] for x in build_info[job_name].values()]
header = "Version:," + ",".join(versions) + '\n'
csv_tables[job_name].append(header)
while not data_queue.empty():
result = data_queue.get()
anomaly_classifications.extend(result["results"])
csv_tables[result["job_name"]].extend(result["csv_table"])
for item in result["logs"]:
if item[0] == "INFO":
logging.info(item[1])
elif item[0] == "ERROR":
logging.error(item[1])
elif item[0] == "DEBUG":
logging.debug(item[1])
elif item[0] == "CRITICAL":
logging.critical(item[1])
elif item[0] == "WARNING":
logging.warning(item[1])
del data_queue
# Terminate all workers
for worker in workers:
worker.terminate()
worker.join()
# Write the tables:
for job_name, csv_table in csv_tables.items():
file_name = spec.cpta["output-file"] + "-" + job_name + "-trending"
with open("{0}.csv".format(file_name), 'w') as file_handler:
file_handler.writelines(csv_table)
txt_table = None
with open("{0}.csv".format(file_name), 'rb') as csv_file:
csv_content = csv.reader(csv_file, delimiter=',', quotechar='"')
line_nr = 0
for row in csv_content:
if txt_table is None:
txt_table = prettytable.PrettyTable(row)
else:
if line_nr > 1:
for idx, item in enumerate(row):
try:
row[idx] = str(round(float(item) / 1000000, 2))
except ValueError:
pass
try:
txt_table.add_row(row)
except Exception as err:
logging.warning("Error occurred while generating TXT "
"table:\n{0}".format(err))
line_nr += 1
txt_table.align["Build Number:"] = "l"
with open("{0}.txt".format(file_name), "w") as txt_file:
txt_file.write(str(txt_table))
# Evaluate result:
if anomaly_classifications:
result = "PASS"
for classification in anomaly_classifications:
if classification == "regression" or classification == "outlier":
result = "FAIL"
break
else:
result = "FAIL"
logging.info("Partial results: {0}".format(anomaly_classifications))
logging.info("Result: {0}".format(result))
return result