mozregression GUI: now available for Linux

Jun 29th, 2020

Mozilla mozregression

Thanks to @AnAverageHuman, mozregression once again has an easy to use and install GUI version for Linux! This used to work a few years ago, but got broken with some changes in the mozregression-python2 era and didn’t get resolved until now:

This is an area where using telemetry in mozregression can help us measure the impact of a change like this: although Windows still dominates in terms of marketshare, Linux is very widely used by contributors — of the usage of mozregression in the past 2 months, fully 30% of the sessions were on Linux (and it is possible we were undercounting that due to bug 1646402):

link to query (internal-only)

It will be interesting to watch the usage numbers for Linux evolve over the next few months. In particular, I’m curious to see what percentage of users on that platform prefer a GUI.

Appendix: reducing mozregression-GUI’s massive size

One thing that’s bothered me a bunch lately is that the mozregression GUI’s size is massive and this is even more apparent on Linux, where the initial distribution of the GUI came in at over 120 megabytes! Why so big? There were a few reasons:

  1. PySide2 (the GUI library we use) is very large (10s of megabytes), and PyInstaller packages all of it by default into your application distribution.
  2. The binary/rust portions of the Glean Python SDK were been built with debugging information included (basically as a carry-over when it was a pre-alpha product), which made it 38 megabytes big (!) on Linux.
  3. On Linux at least, a large number of other system libraries are packaged into the distribution.

A few aspects of this were under our control: Ian Moody (:Kwan) and myself crafted a script to manually remove unneeded PySide2 libraries as part of the packaging process. The Glean team was awesome-as-always and quickly rebuilt Glean without debugging information (this was basically an oversight). Finally, I managed to shave off a few more megabytes by reverting the Linux build to an earlier version of Ubuntu (Xenial), which is something I had been meaning to do anyway.

Even after doing all of these things, the end result is still a little underwhelming: the mozregression GUI distribution on Linux is still 79.5 megabytes big. There are probably other things we could do, but we’re definitely entering the land of diminishing returns.

Honestly, my main takeaway is just not to build an application like this in Python unless you absolutely have to (e.g. you’re building an application which needs system-level access). The web is a pretty wonderful medium for creating graphical applications these days, and by using it you sidestep these type of installation issues.

The humble blog

May 24th, 2020


I’ve been thinking a lot about markdown, presentation, dashboards, and other frontendy sorts of things lately, partly inspired by my work on Iodide last year, partly inspired by my recent efforts on improving I haven’t fully fleshed out my thoughts on this yet, but in general I think blogs (for some value of “blog”) are still a great way to communicate ideas and concepts to an interested audience.

Like many organizations, Mozilla’s gone down the path of Google Docs, Zoom and Slack which makes me more than a little sad: good ideas disappear down the memory hole super quickly with these tools, not to mention the fact that they are closed-by-default (even to people inside Mozilla!). My view on “open” is a bit more nuanced than it used to be: I no longer think everything need be all-public, all-the-time— but I still think talking about and through our ideas (even if imperfectly formed or stated) with a broad audience builds trust and leads to better outcomes.

Is there some way we can blend some of these old school ideas (blogs, newsgroups, open discussion forums) with better technology and social practices? Let’s find out.

A principled reorganization of

May 11th, 2020


(this post is aimed primarily at an internal audience, but I thought I’d go ahead and make it public as a blog post)

I’ve been thinking a bunch over the past few months about the Mozilla data organization’s documentation story. We have a first class data platform here at Mozilla, but using it to answer questions, especially for newer employees, can be quite intimidating. As we continue our collective journey to becoming a modern data-driven organization, part of the formula for unlocking this promise is making the tools and platforms we create accessible to a broad internal audience.

My data peers are a friendly group of people and we have historically been good at answering questions on forums like the #fx-metrics slack channel: we’ll keep doing this. That said, our time is limited: we need a common resource for helping bring people up to speed on how to use the data platform to answer common questions.

Our documentation site,, was meant to be this resource: however in the last couple of years an understanding of its purpose has been (at least partially) lost and it has become somewhat overgrown with content that isn’t very relevant to those it’s intended to help.

This post’s goal is to re-establish a mission for our documentation site — towards the end, some concrete proposals on what to change are also outlined.

Setting the audience was and is meant to be a resource useful for data practitioners within Mozilla.

Examples of different data practioners and their use cases:

There are a range of skills that these different groups bring to the table, but there are some common things we expect a data practitioner to have, whatever their formal job description:

This also excludes a few groups:

What do these users need?

In general, a data practitioner is trying to answer a specific set of questions in the context of an exploration. There are a few things that they need:

What serves this need?

A few years ago, Ryan Harter did an extensive literature review on writing documentation on technical subjects - the take away from this exploration is that the global consensus is that we should focus most of our attention on writing practical tutorials which enables our users to perform specific tasks in the service of the above objective.

There is a proverb, allegedly attributed to Confucius which goes something like this:

“I hear and I forget. I see and I remember. I do and I understand.”

The understanding we want to build is how to use our data systems and tools to answer questions. Some knowledge of how our data platform works is no doubt necessary to accomplish this, but it is mostly functional knowledge we care about imparting to data practitioners: the best way to build this understanding is to guide users in performing tasks.

This makes sense intuitively, but it is also borne out by the data that this is what our users are looking for. Looking through the top pages on Google Analytics, virtually all of them1 refer either to a cookbook or howto guide:

Happily, this allows us to significantly narrow our focus for We no longer need to worry about:

Scanning through the above, you’ll see a common theme: avoid overly detailed reference material. The above is not to say that we should avoid background documentation altogether. For example, an understanding of how our data pipeline works is key to understanding how up-to-date a dashboard is expected to be. However, this type of documentation should be written bearing in mind the audience (focusing on what they need to know as data practitioners) and should be surfaced towards the end of the documentation as supporting material.

As an exception, there is also a very small amount of reference documentation which we want to put at top-level because it is so important: for example the standard metrics page describes how we define “MAU” and “DAU”: these are measures that we want to standardize in the organization, and not have someone re-invent every time they produce an analysis or report. However, we should be very cautious about how much of this “front facing” material we include: if we overwhelm our audience with details right out of the gate, they are apt to ignore them.

Concrete actions


  1. For some reason which I cannot fathom, a broad (non-Mozilla) audience seems unusually interested in our SQL style guide. 

  2. The current Firefox data documentation has a project glossary that is riddled with links to obsolete and unused projects. 

  3. has a huge section devoted to derived datasets (20+), many of which are obsolete or not recommended. At the same time, we are missing explicit reference material for the most commonly used tables in BigQuery (e.g. telemetry.main). 

This Week in Glean: mozregression telemetry (part 2)

May 8th, 2020

Mozilla Telemetry Glean mozregression

(“This Week in Glean” is a series of blog posts that the Glean Team at Mozilla is using to try to communicate better about our work. They could be release notes, documentation, hopes, dreams, or whatever: so long as it is inspired by Glean. You can find an index of all TWiG posts online.)

This is a special guest post by non-Glean-team member William Lachance!

This is a continuation of an exploration of adding Glean-based telemetry to a python application, in this case mozregression, a tool for automatically finding the source of Firefox regressions (breakage).

When we left off last time, we had written some test scripts and verified that the data was visible in the debug viewer.

Adding Telemetry to mozregression itself

In many ways, this is pretty similar to what I did inside the sample application: the only significant difference is that these are shipped inside a Python application that is meant to be be installable via pip. This means we need to specify the pings.yaml and metrics.yaml (located inside the mozregression subirectory) as package data inside

    package_data={"mozregression": ["*.yaml"]},

There were also a number of Glean SDK enhancements which we determined were necessary. Most notably, Michael Droettboom added 32-bit Windows wheels to the Glean SDK, which we need to make building the mozregression GUI on Windows possible. In addition, some minor changes needed to be made to Glean’s behaviour for it to work correctly with a command-line tool like mozregression — for example, Glean used to assume that Telemetry would always be disabled via a GUI action so that it would send a deletion ping, but this would obviously not work in an application like mozregression where there is only a configuration file — so for this case, Glean needed to be modified to check if it had been disabled between runs.

Many thanks to Mike (and others on the Glean team) for so patiently listening to my concerns and modifying Glean accordingly.

Getting Data Review

At Mozilla, we don’t just allow random engineers like myself to start collecting data in a product that we ship (even a semi-internal like mozregression). We have a process, overseen by Data Stewards to make sure the information we gather is actually answering important questions and doesn’t unnecessarily collect personally identifiable information (e.g. email addresses).

You can see the specifics of how this worked out in the case of mozregression in bug 1581647.


Glean has some fantastic utilities for generating markdown-based documentation on what information is being collected, which I have made available on GitHub:

The generation of this documentation is hooked up to mozregression’s continuous integration, so we can sure it’s up to date.

I also added a quick note to mozregression’s web site describing the feature, along with (very importantly) instructions on how to turn it off.

Enabling Data Ingestion

Once a Glean-based project has passed data review, getting our infrastructure to ingest it is pretty straightforward. Normally we would suggest just filing a bug and let us (the data team) handle the details, but since I’m on that team, I’m going to go a (little bit) of detail into how the sausage is made.

Behind the scenes, we have a collection of ETL (extract-transform-load) scripts in the probe-scraper repository which are responsible for parsing the ping and probe metadata files that I added to mozregression in the step above and then automatically creating BigQuery tables and updating our ingestion machinery to insert data passed to us there.

There’s quite a bit of complicated machinery being the scenes to make this all work, but since it’s already in place, adding a new thing like this is relatively simple. The changeset I submitted as part of a pull request to probe-scraper was all of 9 lines long:

diff --git a/repositories.yaml b/repositories.yaml
index dffcccf..6212e55 100644
--- a/repositories.yaml
+++ b/repositories.yaml
@@ -239,3 +239,12 @@ firefox-android-release:
     - org.mozilla.components:browser-engine-gecko-beta
     - org.mozilla.appservices:logins
     - org.mozilla.components:support-migration
+  app_id: org-mozilla-mozregression
+  notification_emails:
+    -
+  url: ''
+  metrics_files:
+    - 'mozregression/metrics.yaml'
+  ping_files:
+    - 'mozregression/pings.yaml'

A Pretty Graph

With the probe scraper change merged and deployed, we can now start querying! A number of tables are automatically created according to the schema outlined above: notably “live” and “stable” tables corresponding to the usage ping. Using we can start exploring what’s out there. Here’s a quick query I wrote up:

SELECT DATE(submission_timestamp) AS date,
       metrics.string.usage_variant AS variant,
FROM `moz-fx-data-shared-prod`.org_mozilla_mozregression_stable.usage_v1
WHERE DATE(submission_timestamp) >= '2020-04-14'
  AND client_info.app_display_version NOT LIKE ''
GROUP BY date, variant;

… which generates a chart like this:

This chart represents the absolute volume of mozregression usage since April 14th 2020 (around the time when we first released a version of mozregression with Glean telemetry), grouped by mozregression “variant” (GUI, console, and mach) and date - you can see that (unsurprisingly?) the GUI has the highest usage. I’ll talk about this more in an upcoming installment, speaking of…

Next Steps

We’re not done yet! Next time, we’ll look into making a public-facing dashboard demonstrating these results and making an aggregated version of the mozregression telemetry data publicly accessible to researchers and the general public. If we’re lucky, there might even be a bit of data science. Stay tuned!

mozregression for MacOS

Apr 24th, 2020

Mozilla mozregression

Just a quick note that, as a side-effect of the work I mentioned a while ago to add telemetry to mozregression, mozregression now has a graphical Mac client! It’s a bit of a pain to install (since it’s unsigned), but likely worlds easier for the average person to get going than the command-line version. Please feel free to point people to it if you’re looking to get a regression range for a MacOS-specific problem with Firefox.

More details: The Glean Python SDK, which mozregression now uses for telemetry, requires Python 3. This provided the impetus to port the GUI itself to Python 3 and PySide2 (the modern incarnation of PyQt), which brought with it a much easier installation/development experience for the GUI on platforms like Mac and Linux.

I haven’t gotten around to producing GUI binaries for the Linux yet, but it should not be much work.

Speaking of Glean, mozregression, and Telemetry, stay tuned for more updates on that soon. It’s been an adventure!

This week in Glean (special guest post): mozregression telemetry (part 1)

Feb 28th, 2020

Mozilla Telemetry Glean mozregression

(“This Week in Glean” is a series of blog posts that the Glean Team at Mozilla is using to try to communicate better about our work. They could be release notes, documentation, hopes, dreams, or whatever: so long as it is inspired by Glean. You can find an index of all TWiG posts online.)

This is a special guest post by non-Glean-team member William Lachance!

As I mentioned last time I talked about mozregression, I have been thinking about adding some telemetry to the system to better understand the usage of this tool, to justify some part of Mozilla spending some cycles maintaining and improving it (assuming my intuition that this tool is heavily used is confirmed).

Coincidentally, the Telemetry client team has been working on a new library for measuring these types of things in a principled way called Glean, which even has python bindings! Using this has the potential in saving a lot of work: not only does Glean provide a framework for submitting data, our backend systems are automatically set up to process data submitted via into Glean into BigQuery tables, which can then easily be queried using tools like

I thought it might be useful to go through some of what I’ve been exploring, in case others at Mozilla are interested in instrumenting their pet internal tools or projects. If this effort is successful, I’ll distill these notes into a tutorial in the Glean documentation.

Initial steps: defining pings and metrics

The initial step in setting up a Glean project of any type is to define explicitly the types of pings and metrics. You can look at a “ping” as being a small bucket of data submitted by a piece of software in the field. A “metric” is something we’re measuring and including in a ping.

Most of the Glean documentation focuses on browser-based use-cases where we might want to sample lots of different things on an ongoing basis, but for mozregression our needs are considerably simpler: we just want to know when someone has used it along with a small number of non-personally identifiable characteristics of their usage, e.g. the mozregression version number and the name of the application they are bisecting.

Glean has the concept of event pings, but it seems like those are there more for a fine-grained view of what’s going on during an application’s use. So let’s define a new ping just for ourselves, giving it the unimaginative name “usage”. This goes in a file called pings.yaml:

$schema: moz://

  description: >
    A ping to record usage of mozregression
  include_client_id: true

We also need to define a list of things we want to measure. To start with, let’s just test with one piece of sample information: the app we’re bisecting (e.g. “Firefox” or “Gecko View Example”). This goes in a file called metrics.yaml:

$schema: moz://

    type: string
    description: >
      The name of the app being bisected
    expires: never
      - usage

The data_reviews sections in both of the above are obviously bogus, we will need to actually get data review before landing and using this code, to make sure that we’re in conformance with Mozilla’s data collection policies.

Testing it out

But in the mean time, we can test our setup with the Glean debug pings viewer by setting a special tag (mozregression-test-tag) on our output. Here’s a small python script which does just that:

from pathlib import Path
from glean import Glean, Configuration
from glean import (load_metrics,

mozregression_path = Path.home() / '.mozilla2' / 'mozregression'

    data_dir=mozregression_path / "data"

pings = load_pings("pings.yaml")
metrics = load_metrics("metrics.yaml")"reality")

Running this script on my laptop, I see that a respectable JSON payload was delivered to and processed by our servers:

As you can see, we’re successfully processing both the “version” number of mozregression, some characteristics of the machine sending the information (my MacBook in this case), as well as our single measure. We also have a client id, which should tell us roughly how many distinct installations of mozregression are sending pings. This should be more than sufficient for an initial “mozregression usage dashboard”.

Next steps

There are a bunch of things I still need to work through before landing this inside mozregression itself. Notably, the Glean python bindings are python3-only, so we’ll need to port the mozregression GUI to python 3 before we can start measuring usage there. But I’m excited at how quickly this work is coming together: stay tuned for part 2 in a few weeks.

Conda is pretty great

Jan 13th, 2020

Mozilla Telemetry

Lately the data engineering team has been looking into productionizing (i.e. running in Airflow) a bunch of models that the data science team has been producing. This often involves languages and environments that are a bit outside of our comfort zone — for example, the next version of Mission Control relies on the R-stan library to produce a model of expected crash behaviour as Firefox is released.

To make things as simple and deterministic as possible, we’ve been building up Docker containers to run/execute this code along with their dependencies, which makes things nice and reproducible. My initial thought was to use just the language-native toolchains to build up my container for the above project, but quickly found a number of problems:

  1. For local testing, Docker on Mac is slow: when doing a large number of statistical calculations (as above), you can count on your testing iterations taking 3 to 4 (or more) times longer.
  2. On initial setup, the default R packaging strategy is to have the user of a package like R-stan recompile from source. This can take forever if you have a long list of dependencies with C-compiled extensions (pretty much a given if you’re working in the data space): rebuilding my initial docker environment for missioncontrol-v2 took almost an hour. This isn’t just a problem for local development: it also makes continuous integration using a service like Circle or Travis expensive and painful.

I had been vaguely aware of Conda for a few years, but didn’t really understand its value proposition until I started working on the above project: why bother with a heavyweight package manager when you already have Docker to virtualize things? The answer is that it solves both of the above problems: for local development, you can get something more-or-less identical to what you’re running inside Docker with no performance penalty whatsoever. And for building the docker container itself, Conda’s package repository contains pre-compiled versions of all the dependencies you’d want to use for something like this (even somewhat esoteric libraries like R-stan are available on conda-forge), which brought my build cycle times down to less than 5 minutes.

tl;dr: If you have a bunch of R / python code you want to run in a reproducible manner, consider Conda.

Using BigQuery JavaScript UDFs to analyze Firefox telemetry for fun & profit

Oct 30th, 2019

Mozilla Telemetry

For the last year, we’ve been gradually migrating our backend Telemetry systems from AWS to GCP. I’ve been helping out here and there with this effort, most recently porting a job we used to detect slow tab spinners in Firefox nightly, which produced a small dataset that feeds a small adhoc dashboard which Mike Conley maintains. This was a relatively small task as things go, but it highlighted some features and improvements which I think might be broadly interesting, so I decided to write up a small blog post about it.

Essentially all this dashboard tells you is what percentage of the Firefox nightly population saw a tab spinner over the past 6 months. And of those that did see a tab spinner, what was the severity? Essentially we’re just trying to make sure that there are no major regressions of user experience (and also that efforts to improve things bore fruit):

Pretty simple stuff, but getting the data necessary to produce this kind of dashboard used to be anything but trivial: while some common business/product questions could be answered by a quick query to clients_daily, getting engineering-specific metrics like this usually involved trawling through gigabytes of raw heka encoded blobs using an Apache Spark cluster and then extracting the relevant information out of the telemetry probe histograms (in this case, FX_TAB_SWITCH_SPINNER_VISIBLE_MS and FX_TAB_SWITCH_SPINNER_VISIBLE_LONG_MS) contained therein.

The code itself was rather complicated (take a look, if you dare) but even worse, running it could get very expensive. We had a 14 node cluster churning through this script daily, and it took on average about an hour and a half to run! I don’t have the exact cost figures on hand (and am not sure if I’d be authorized to share them if I did), but based on a back of the envelope sketch, this one single script was probably costing us somewhere on the order of $10-$40 a day (that works out to between $3650-$14600 a year).

With our move to BigQuery, things get a lot simpler! Thanks to the combined effort of my team and data operations[1], we now produce “stable” ping tables on a daily basis with all the relevant histogram data (stored as JSON blobs), queryable using relatively vanilla SQL. In this case, the data we care about is in telemetry.main (named after the main ping, appropriately enough). With the help of a small JavaScript UDF function, all of this data can easily be extracted into a table inside a single SQL query scheduled by

  udf_js_json_extract_highest_long_spinner (input STRING)
  LANGUAGE js AS """
    if (input == null) {
      return 0;
    var result = JSON.parse(input);
    var valuesMap = result.values;
    var highest = 0;
    for (var key in valuesMap) {
      var range = parseInt(key);
      if (valuesMap[key]) {
        highest = range > 0 ? range : 1;
    return highest;

SELECT build_id,
sum (case when highest >= 64000 then 1 else 0 end) as v_64000ms_or_higher,
sum (case when highest >= 27856 and highest < 64000 then 1 else 0 end) as v_27856ms_to_63999ms,
sum (case when highest >= 12124 and highest < 27856 then 1 else 0 end) as v_12124ms_to_27855ms,
sum (case when highest >= 5277 and highest < 12124 then 1 else 0 end) as v_5277ms_to_12123ms,
sum (case when highest >= 2297 and highest < 5277 then 1 else 0 end) as v_2297ms_to_5276ms,
sum (case when highest >= 1000 and highest < 2297 then 1 else 0 end) as v_1000ms_to_2296ms,
sum (case when highest > 0 and highest < 50 then 1 else 0 end) as v_0ms_to_49ms,
sum (case when highest >= 50 and highest < 100 then 1 else 0 end) as v_50ms_to_99ms,
sum (case when highest >= 100 and highest < 200 then 1 else 0 end) as v_100ms_to_199ms,
sum (case when highest >= 200 and highest < 400 then 1 else 0 end) as v_200ms_to_399ms,
sum (case when highest >= 400 and highest < 800 then 1 else 0 end) as v_400ms_to_799ms,
count(*) as count
(select build_id, client_id, max(greatest(highest_long, highest_short)) as highest
    SUBSTR(application.build_id, 0, 8) as build_id,
    udf_js_json_extract_highest_long_spinner(payload.histograms.FX_TAB_SWITCH_SPINNER_VISIBLE_LONG_MS) AS highest_long,
    udf_js_json_extract_highest_long_spinner(payload.histograms.FX_TAB_SWITCH_SPINNER_VISIBLE_MS) as highest_short
FROM telemetry.main
    AND normalized_os='Windows'
    AND application.build_id > FORMAT_DATE("%Y%m%d", DATE_SUB(CURRENT_DATE(), INTERVAL 2 QUARTER))
    AND DATE(submission_timestamp) >= DATE_SUB(CURRENT_DATE(), INTERVAL 2 QUARTER))
group by build_id, client_id) group by build_id;

In addition to being much simpler, this new job is also way cheaper. The last run of it scanned just over 1 TB of data, meaning it cost us just over $5. Not as cheap as I might like, but considerably less expensive than before: I’ve also scheduled it to only run once every other day, since Mike tells me he doesn’t need this data any more often than that.

[1] I understand that Jeff Klukas, Frank Bertsch, Daniel Thorn, Anthony Miyaguchi, and Wesley Dawson are the principals involved - apologies if I’m forgetting someone.

Metrics Graphics: Stepping back for a while

Sep 26th, 2019

Metrics Graphics Mozilla

Just a note that I’ve decided to step back from metrics graphics maintenance for the time being, which means that the project is essentially unowned. This has sort of been the case for a while, but I figured I should probably make it official.

If you follow the link to the metrics graphics repository, you’ll note that the version has been bumped to “3.0-alpha3”. I was this close to making one last new release this afternoon but decided I didn’t want to potentially break existing users who were fine using the last “official” version (v3.0 bumps the version of d3 used to “5”, among other breaking changes). I’d encourage people who want to continue using the library to make a fork and publish a copy under their user or organization name on npm.

mozregression update: python 3 edition

Sep 16th, 2019

Mozilla mozregression

For those who are still wondering, yup, I am still maintaining mozregression, though increasingly reluctantly. Given how important this project is to the development of Firefox (getting a regression window using mozregression is standard operating procedure whenever a new bug is reported in Firefox), it feels like this project is pretty vital, so I continue out of some sense of obligation — but really, someone more interested in Mozilla’a build, automation and testing systems would be better suited to this task: over the past few years, my interests/focus have shifted away from this area to building up Mozilla’s data storage and visualization platform.

This post will describe some of the things that have happened in the last year and where I see the project going. My hope is to attract some new blood to add some needed features to the project and maybe take on some of the maintainership duties.

python 3

The most important update is that, as of today, the command-line version of mozregression (v3.0.1) should work with python 3.5+. modernize did most of the work for us, though there were some unit tests that needed updating: special thanks to @gloomy-ghost for helping with that.

For now, we will continue to support python 2.7 in parallel, mainly because the GUI has not yet been ported to python 3 (more on that later) and we have CI to make sure it doesn’t break.

other updates

The last year has mostly been one of maintenance. Thanks in particular to Ian Moody (:kwan) for his work throughout the year — including patches to adapt mozregression support to our new updates policy and shippable builds (bug 1532412), and Kartikaya Gupta (:kats) for adding support for bisecting the GeckoView example app (bug 1507225).

future work

There are a bunch of things I see us wanting to add or change with mozregression over the next year or so. I might get to some of these if I have some spare cycles, but probably best not to count on it:

If you’re interested in working on any of the above, please feel free to dive in on one of the above bugs. I can’t offer formal mentorship, but am happy to help out where I can.