subprocess_crash_submit

subprocess_crash_submit.ipynb

subprocess_crash_submit.py

# coding: utf-8

# ### Subprocess Crash Submission Rates

# In[1]:

import ujson as json
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
import plotly.plotly as py
from plotly.graph_objs import *

from moztelemetry import get_pings, get_pings_properties, get_one_ping_per_client, get_clients_history, get_records

get_ipython().magic(u'pylab inline')


# In[2]:

sc.defaultParallelism


# In[43]:

pings = get_pings(sc, app="Firefox", submission_date="20161101", fraction=0.4)


# In[44]:

subset = get_pings_properties(pings, ["clientId",
                                      "application/channel",
                                      "payload/keyedHistograms/PROCESS_CRASH_SUBMIT_SUCCESS/content-crash",
                                      "payload/keyedHistograms/PROCESS_CRASH_SUBMIT_SUCCESS/plugin-crash",
                                      "payload/keyedHistograms/SUBPROCESS_CRASHES_WITH_DUMP/content",
                                      "payload/keyedHistograms/SUBPROCESS_CRASHES_WITH_DUMP/plugin",
                                     ])


# In[45]:

subset = subset.filter(lambda p: p["application/channel"] is not None and
                      (p["payload/keyedHistograms/PROCESS_CRASH_SUBMIT_SUCCESS/content-crash"] is not None or
                      p["payload/keyedHistograms/PROCESS_CRASH_SUBMIT_SUCCESS/plugin-crash"] is not None or
                      p["payload/keyedHistograms/SUBPROCESS_CRASHES_WITH_DUMP/content"] is not None or
                      p["payload/keyedHistograms/SUBPROCESS_CRASHES_WITH_DUMP/plugin"] is not None))


# In[46]:

def pair_or_none(p, prefix, hgram, crash):
    return (prefix + crash, 1) if p[hgram + "/" + crash] is not None else None


def gimme_count(p, hgram, crashes):
    prefix = p["application/channel"] + "-"
    return [pair_or_none(p, prefix, hgram, crash) for crash in crashes]

crash_counts = subset                .flatMap(lambda p: gimme_count(p, "payload/keyedHistograms/SUBPROCESS_CRASHES_WITH_DUMP", ["content", "plugin"]))                .filter(lambda pair: pair is not None)                .countByKey()
crash_counts


# In[47]:

crash_reports = subset                    .flatMap(lambda p: gimme_count(p, "payload/keyedHistograms/PROCESS_CRASH_SUBMIT_SUCCESS", ["content-crash", "plugin-crash"]))                    .filter(lambda pair: pair is not None)                    .countByKey()
crash_reports


# In[48]:

{k: (1.0 * v / crash_counts.get(k[:-6], v)) for (k, v) in crash_reports.iteritems()}