Docker is a pretty nifty piece of software. It's essentially a glorified wrapper around Linux Containers. But, calling it that is doing it an injustice.
Docker interests me because it allows simple environment isolation and repeatability. I can create a run-time environment once, package it up, then run it again on any other machine. Furthermore, everything that runs in that environment is isolated from the underlying host (much like a virtual machine). And best of all, everything is fast and simple.
For my initial experimentation with Docker, I decided to create an environment for building Firefox.
To build Firefox with Docker, you'll first need to install Docker. That's pretty simple.
Then, it's just a matter of creating a new container with our build environment:
curl https://gist.github.com/indygreg/5608534/raw/30704c59364ce7a8c69a02ee7f1cfb23d1ffcb2c/Dockerfile | docker build
The output will look something like:
FROM ubuntu:12.10 MAINTAINER Gregory Szorc "gps@mozilla.com" RUN apt-get update ===> d2f4faba3834 RUN dpkg-divert --local --rename --add /sbin/initctl && ln -s /bin/true /sbin/initctl ===> aff37cc837d8 RUN apt-get install -y autoconf2.13 build-essential unzip yasm zip ===> d0fc534feeee RUN apt-get install -y libasound2-dev libcurl4-openssl-dev libdbus-1-dev libdbus-glib-1-dev libgtk2.0-dev libiw-dev libnotify-dev libxt-dev mesa-common-dev uuid-dev ===> 7c14cf7af304 RUN apt-get install -y binutils-gold ===> 772002841449 RUN apt-get install -y bash-completion curl emacs git man-db python-dev python-pip vim ===> 213b117b0ff2 RUN pip install mercurial ===> d3987051be44 RUN useradd -m firefox ===> ce05a44dc17e Build finished. image id: ce05a44dc17e ce05a44dc17e
As you can see, it is essentially bootstrapping an environment to build Firefox.
When this has completed, you can activate a shell in the container by taking the image id printed at the end and running it:
docker run -i -t ce05a44dc17e /bin/bash # You should now be inside the container as root. su - firefox hg clone https://hg.mozilla.org/mozilla-central cd mozilla-central ./mach build
If you want to package up this container for distribution, you just find its ID then export it to a tar archive:
docker ps -a # Find ID of container you wish to export. docker export 2f6e0edf64e8 > image.tar # Distribute that file somewhere. docker import - < image.tar
Simple, isn't it?
I think it would be rad if Release Engineering used Docker for managing their Linux builder configurations. Want to develop against the exact system configuration that Mozilla uses in its automation - you could do that. No need to worry about custom apt repositories, downloading custom toolchains, keeping everything isolated from the rest of your system, etc: Docker does that all automatically. Mozilla simply needs to publish Docker images on the Internet and anybody can come along and reproduce the official environment with minimal effort. Once we do that, there are few excuses for someone breaking Linux builds because of an environment discrepancy.
Release Engineering could also use Docker to manage isolation of environments between builds. For example, it could spin up a new container for each build or test job. It could even save images from the results of these jobs. Have a weird build failure like a segmentation fault in the compiler? Publish the Docker image and have someone take a look! No need to take the builder offline while someone SSH's into it. No need to worry about the probing changing state because you can always revert to the state at the time of the failure! And, builds would likely start faster. As it stands, our automation spends minutes managing packages before builds begin. This lag would largely be eliminated with Docker. If nothing else, executing automation jobs inside a container would allow us to extract accurate resource usage info (CPU, memory, I/O) since the Linux kernel effectively gives containers their own namespace independent of the global system's.
I might also explore publishing Docker images that construct an ideal development environment (since getting recommended tools in the hands of everybody is a hard problem).
Maybe I'll even consider hooking up build system glue to automatically run builds inside containers.
Lots of potential here.
I encourage Linux users to play around with Docker. It enables some new and exciting workflows and is a really powerful tool despite its simplicity. So far, the only major faults I have with it are that the docs say it should not be used in production (yet) and it only works on Linux.
]]>Kyle Huey released version 1.7 of our Windows build environment. It contains a newer version of Python and a modern version of Mercurial among other features.
I highly recommend every Windows developer update ASAP. Please note that you will likely encounter Python errors unless you clobber your build.
I used my power as module owner to create a submodule of the build config module whose scope is the (largely mechanical) transition of content from Makefile.in to moz.build files. I granted Joey Armstrong and Mike Shal peer status for this module. I would like to eventually see both elevated to build peers of the main build module.
The following progress has been made:
In bug 850380 I'm trying to land non-recursive building of XPIDL files. As part of this I'm trying to combine the generation of .xpt and .h for each input .idl file into a single process call because profiling revealed that parsing the IDL consumes most of the CPU time. This shaves a few dozen seconds off of build times.
I have encounterd multiple pymake bugs when developing this patch, which is the primary reason it hasn't landed yet.
I was looking at my build logs and noticed WebIDL generation was taking longer than I thought it should. I filed bug 861587 to investigate making it faster. While my initial profiling turned out to be wrong, Boris Zbarsky looked into things and discovered that the serialization and deserialization of the parser output was extremely slow. He is currently trying to land a refactor of how WebIDL bindings are handled. The early results look very promising.
I think the bug is a good example of the challenges we face improving the build system, as Boris can surely attest.
Joel Maher is injecting sanity into the naming scheme of test directories in bug 852065.
Jeff Hammel, Joel Maher, Ted Mielczarek, and I are working out using manifests for mochitests (like xpcshell tests) in bug 852416.
I extracted the mach core to a standalone repository and added it to PyPI.
Mach now categorizes commands in its help output.
Now that the Windows build environment ships with Python 2.7.4, I've filed bug 870420 to require Python 2.7.3+ to build the tree. We already require Python 2.7.0+. I want to bump the point release because there are many small bug fixes in 2.7.3, especially around Python 3 compatibility.
This is currently blocked on RelEng rolling out 2.7.3 to all the builders.
Now that xpcshell test manifests are defined in moz.build files, we theoretically don't need the master manifest. Joshua Cranmer is working on removing them in bug 869635.
Benoit Gerard and Mike Hommey are working in bug 844288 to dual link libxul so GTests can eventually be enabled and executed as part of our automation.
This will regress build times since we need to link libxul twice. But, giving C++ developers the ability to write unit tests with a real testing framework is worth it, in my opinion.
ICU was briefly enabled in bug 853301 but then backed out because it broke cross-compiling. It should be on track for enabling in Firefox 24.
I gave mozbase a class to record system resource usage. I plan to eventually hook this up to the build system so the build system records how long it took to perform key events. This will give us better insight into slow and inefficient parts of the build and will help us track build system speed improvements over time.
I'm working on requiring lists in moz.build be sorted. Work is happening in bug 863069.
This idea started as a suggestion on the dev-platform list. If anyone has more great ideas, don't hold them back!
Nicholas Alexander taught mach how to build intelligently by importing some of Josh Matthews' smartmake tool's functionality into the tree.
Kyle Huey and Ted Mielczarek collaborated to fix the source server.
Auto clobber functionality was added to the tree. After flirting briefly with on-by-default, we changed it to opt-in. When you encounter it, it will tell you how to enable it.
I was looking at build logs and identified we were inefficiently performing clobber.
Massimo Gervasini and Chris AtLee deployed changes to automation to make it more efficient. My measurements showed a Windows try build that took 15 fewer minutes to start - a huge improvement.
RelEng is tracking the global deployment of Mercurial 2.5.4. hg.mozilla.org is currently running 2.0.2 and automation is all over the map. The upgrade should make Mercurial operations faster and more robust across the board.
I'm considering adding code to mach or the build system that prompts the user when her Mercurial is out of date (since an out of date Mercurial can result in a sub-par user experience).
Nathan Froyd is leading an effort to parallelize reftest execution. If he pulls this off, it could shave hours off of the total automation load per checkin. Go Nathan!
I am mentoring a pair of interns this summer. I'm still working out the final set of goals, but I'm keen to have one of them overhaul the MozillaBuild Windows development environment. Cross your fingers.
]]>I believe that Mozilla's build system has had a poor developer experience for as long as there has been a Mozilla build system. Getting started with Firefox development was a rite of passage. It required following (often out-of-date) directions on MDN. It required finding pages through MDN search or asking other people for info over IRC. It was the kind of process that turned away potential contributors because it was just too damn hard.
mach - while born out of my initial efforts to radically change the build system proper - morphed into a generic command dispatching framework by the time it landed in mozilla-central. It has one overarching purpose: provide a single gateway point for performing common developer tasks (such as building the tree and running tests). The concept was nothing new - individual developers had long coded up scripts and tools to streamline workflows. Some even published these for others to use. What set mach apart was a unified interface for these commands (the mach script in the top directory of a checkout) and that these productivity gains were in the tree and thus easily discoverable and usable by everybody without significant effort (just run mach help).
While mach doesn't yet satisfy everyone's needs, it's slowly growing new features and making developers' lives easier with every one. All of this is happening despite that there is not a single person tasked with working on mach full time. Until a few months ago, mach was largely my work. Recently, Matt Brubeck has been contributing a flurry of enhancements - thanks Matt! Ehsan Akhgari and Nicholas Alexander have contributed a few commands as well! There are also a few people with a single command to their name. This is fulfilling my original vision of facilitating developers to scratch their own itches by contributing mach commands.
I've noticed more people referencing mach in IRC channels. And, more people get angry when a mach command breaks or changes behavior. So, I consider the mach experiment a success. Is it perfect, no. If it's not good enough for you, please file a bug and/or code up a patch. If nothing else, please tell me: I love to know about everyone's subtle requirements so I can keep them in mind when refactoring the build system and hacking on mach.
One of the ideas I'm trying to advance is that the object directory should be considered a black box for the majority of developers. In my ideal world, developers don't need to look inside the object directory. Instead, they interact with it through condoned and supported tools (like mach).
I say this for a few reasons. First, as the build config module owner I would like the ability to massively refactor the internals of the object directory without disrupting workflows. If people are interacting directly with the object directory, I get significant push back if things change. This inevitably holds back much-needed improvements and triggers resentment towards me, build peers, and the build system. Not a good situation. Whereas if people are indirectly interacting with the object directory, we simply need to maintain a consistent interface (like mach) and nobody should care if things change.
Second, I believe that the methods used when directly interacting with the object directory are often sub-par compared with going through a more intelligent tool and that productivity suffers as a result. For example, when you type make in inside the object directory you need to know to pass -j8, use make vs pymake, and that you also need to build toolkit/library, etc. Also, by invoking make directly, you bypass other handy features, such as automatic compiler warning aggregation (which only happens if you invoke the build system through mach). If you go through a tool like mach, you should automatically get the most ideal experience possible.
In order for this vision to be realized, we need massive improvements to tools like mach to cover the missing workflows that still require direct object directory interaction. We also need people to start using mach. I think increased mach usage comes after mach has established itself as obviously superior to the alternatives (I already believe it offers this for tasks like running tests).
Nobody likes when they are forced to change a process that has been familiar for years. Developers especially. I get it. That's why I've always attempted to position mach as an alternative to existing workflows. If you don't like mach, you can always fall back to the previous workflow. Or, you can improve mach (patches more than welcome!). Having gone down the please-use-this-tool-it's-better road before at other organizations, I strongly believe that the best method to incur adoption of a new tool is to gradually sway people through obvious superiority and praise (as opposed to a mandate to switch). I've been trying this approach with mach.
Lately, more and more people have been saying things like we should have the build infrastructure build through mach instead of client.mk and why do we need testsuite-targets.mk when we have mach commands. While I personally feel that client.mk and testsuite-targets.mk are antiquated as a developer-facing interface compared to mach, I'm reluctant to eliminate them because I don't like forcing change on others. That being said, there are compelling reasons to eliminate or at least refactor how they work.
Let's take testsuite-targets.mk as an example. This is the make file that provides the targets to run tests (like make xpcshell-test and make mochitest-browser-chrome). What's interesting about this file is that it's only used in local builds: our automation infrastructure does not use testsuite-targets.mk! Instead, mozharness and the old buildbot configs manually build up the command used to invoke the test harnesses. Initially, the mach commands for running tests simply invoked make targets defined in testsuite-targets.mk. Lately, we've been converting the mach commands to invoke the Python test runners directly. I'd argue that the logic for invoke the test runner only needs to live in one place in the tree. Furthermore as a build module peer, I have little desire to support multiple implementations. Especially considering how fragile they can be.
I think we're trending towards an outcome where mach (or the code behind mach commands) transitions into the authoratitive invocation method and legacy interfaces like client.mk and testsuite-targets.mk are reimplemented to either call mach commands or the same routine that powers them. Hopefully this will be completely transparent to developers.
mozconfig files are shell scripts used to define variables consumed by the build system. They are the only officially supported mechanism for configuring how the build system works.
I'd argue mozconfig files are a mediocre solution at best. First, there's the issue of mozconfig statements that don't actually do anything. I've seen no-op mozconfig content cargo culted into the in-tree mozconfigs (used for the builder configurations)! Oops. Second, doing things in mozconfig files is just awkward. Defining the object directory requires mk_add_options MOZ_OBJDIR=some-path. What's mk_add_options? If some-path is relative, what is it relative to? While certainly addressable, the documentation on how mozconfig files work is not terrific and fails to explain many pitfalls. Even with proper documentation, there's still the issue of the file format allowing no-op variable assignments to persist.
I'm very tempted to reinvent build configuration as something not mozconfigs. What exactly, I don't know. mach has support for ini-like configuration files. We could certainly have mach and the build system pull configs from the same file.
I'm not sure what's going to happen here. But deprecating mozconfig files as they are today is part of many of the options.
A lot of developers only have a single mozconfig file (per source tree at least). For these developers, life is easy. You simply install your mozconfig in one of the default locations and it's automagically used when you use mach or client.mk. Easy peasy.
I'm not sure what the relative numbers are, but many developers maintain multiple mozconfig files per source tree. e.g. they'll have one mozconfig to build desktop Firefox and another one for Android. They may have debug variations of each.
Some developers even have a single mozconfig file but leverage the fact that mozconfig files are shell scripts and have their mozconfig dynamically do things depending on the current working directory, value of an environment variable, etc.
I've also seen wrapper scripts that glorify setting environment variables, changing directory, etc and invoke a command.
I've been thinking a lot about providing a common and well-supported solution for switching between active build configurations. Installing mach on $PATH goes a long way to facilitate this. If you are in an object directory, the mozconfig used when that object directory was created is automatically applied. Simple enough. However, I want people to start treating object directories as black boxes. So, I'd rather not see people have their shell inside the object directory.
Whenever I think about solutions, I keep arriving at a virtualenv-like solution. Developers would potentially need to activate a Mozilla build environment (similar to how Windows developers need to launch MozillaBuild). Inside this environment, the shell prompt would contain the name of the current build configuration. Users could switch between configurations using mach switch or some other magic command on the $PATH.
Truth be told, I'm skeptical if people would find this useful. I'm not sure it's that much better than exporting the MOZCONFIG environment variable to define the active config. This one requires more thought.
We use Python extensively in the build system and for common developer tasks. mach is written in Python. moz.build processing is implemented in Python. Most of the test harnesses are written in Python.
Doing practically anything in the tree requires a Python interpreter that knows about all the Python code in the tree and how to load it.
Currently, we have two very similar Python environments. One is a virtualenv created while running configure at the beginning of a build. The other is essentially a cheap knock-off that mach creates when it is launched.
At some point I'd like to consolidate these Python environments. From any Python process we should have a way to automatically bootstrap/activate into a well-defined Python environment. This certainly sounds like establishing a unified Python virtualenv used by both the build system and mach.
Unfortunately, things aren't straightforward. The virtualenv today is constructed in the object directory. How do we determine the current object directory? By loading the mozconfig file. How do we do that? Well, if you are mach, we use Python. And, how does mach know where to find the code to load the mozconfig file? You can see the dilemma here.
A related issue is that of portable build environments. Currently, a lot of our automation recreates the build system's virtualenv from its own configuration (not that from the source tree). This has and will continue to bite us. We'd really like to package up the virtualenv (or at least its config) with tests so there is no potential for discrepancy.
The inner workings of how we integrate with Python should be invisible to most developers. But, I figured I'd capture it here because it's an annoying problem. And, it's also related to an activated build environment. What if we required all developers to activate their shell with a Mozilla build environment (like we do on Windows)? Not only would this solve Python issues, but it would also facilitate simpler config switching (outlined above). Hmmm...
Ever since there was a build system developers have been typing make (or make.py) to build the tree. One of the goals of the transition to moz.build files is to facilitate building the tree with Tup. make will do nothing when you're not using Makefiles! Another goal of the moz.build transition is to start derecursifying the make build system such that we build things in parallel. It's likely we'll produce monolithic make files and then process all targets for a related class IDLs, C++ compilation, etc in one invocation of make. So, uh, what happens during a partial tree build? If a .cpp file from /dom/src/storage is being handled by a monolithic make file invoked by the Makefile at the top of the tree, how does a partial tree build pick that up? Does it build just that target or every target in the monolithic/non-recursive make file?
Unless the build peers go out of our way to install redundant targets in leaf Makefiles, directly invoking make from a subdirectory of the tree won't do what it's done for years.
As I said above, I'm sympathetic to forced changes in procedure, so it's likely we'll provide backwards-compatibile behavior. But, I'd prefer to not do it. I'd first prefer partial-tree builds are not necessary and a full tree build finishes quickly. But, we're not going to get there for a bit. As an alternative, I'll take people building through mach build. That way, we have an easily extensible interface on which to build partial tree logic. We saw this recently when dumbmake/smartmake landed. And, going through mach also reinforces my ideal that the object directory is a black box.
Currently, most state as it pertains to a checkout or build is in the object directory. This is fine for artifacts from the build system. However, there is a whole class of state that arguably shouldn't be in the object directory. Specifically, it shouldn't be clobbered when you rebuild. This includes logs from previous builds, the warnings database, previously failing tests, etc. The list is only going to grow over time.
I'd like to establish a location for semi-persistant state related to the tree and builds. Perhaps we change the clobber logic to ignore a specific directory. Perhaps we start storing things in the user's home directory. Perhaps we could establish a second object directory named the state directory? How would this interact with build environments?
This will probably sit on the backburner until there is a compelling use case for it.
Compiling C++ consumes the bulk of our build time. Anything we can do to speed up C++ compilation will work wonders for our build times.
I'm optimistic things like precompiled headers and compiling multiple .cpp files with a single process invocation will drastically decrease build times. However, no matter how much work we put in to make C++ compilation faster, we still have a giant issue: dependency hell.
As shown in my build system presentation a few months back, we have dozens of header files included by hundreds if not thousands of C++ files. If you change one file: you invalidate build dependencies and trigger a rebuild. This is why whenever files like mozilla-config.h change you are essentially confronted with a full rebuild. ccache may help if you are lucky. But, I fear that as long as headers proliferate the way they do, there is little the build system by itself can do.
My attitude towards this is to wait and see what we can get out of precompiled headers and the like. Maybe that makes it good enough. If not, I'll likely be making a lot of noise at Platform meetings requesting that C++ gurus brainstorm on a solution for reducing header proliferation.
Belive it or not, these are only some of the topics floating around in my head! But I've probably managed to bore everyone enough so I'll call it a day.
I'm always interested in opinions and ideas, especially if they are different from mine. I encourage you to leave a comment if you have something to say.
]]>Here's why.
Our Mozilla server, hg.mozilla.org, is currently running Mercurial 2.0.2. In terms of Mercurial features, stability, and performance, we are light years behind.
You know that annoying phases configuration you need to set when pushing to Try? That's because the server isn't new enough to tell the client the same thing the configuration option does. It will be fixed when the server is upgraded to 2.1+.
Furthermore, we are running the server over NFS, which introduces known badness, including slowness.
I believe we blame Mercurial for issues that would go away if we configured the Mercurial server properly.
Fortunately, it appears the upgrade to 2.5 is near and I've heard we're moving from NFS to local disk storage as part of that. This should go a long way to making the server better. The upgrade can't happen soon enough.
I think a lot of people are ignorant on the features and abilities of Mercurial.
I commonly hear people are dissatisfied with the behavior of their Mercurial client. They encounter performance issues, bugs, corruption, etc. Nine times out of ten this is due to running an old Mercurial release. Just last Friday someone on my team asked me about weird behavior involving file case. My first question: what version of Mercurial are you using? He was running 2.0.2. I told him to upgrade to 2.5+. It fixed his problem. If you aren't running Mercurial 2.5 or newer, upgrade immediately.
I've heard people say we should switch to Git because Git has feature X. Most of the time, Mercurial has these features. Unfortunately, people just don't realize it. When I point them at Mercurial's extensions list their eyes light up and they thank me for making their lives easier.
I think a problem is a lot of new Mozilla contributors knew Git before and only pick up the bare essentials of Mercurial that allow them to land patches. They prefer Git because it is familiar and just don't bother to pick up Mercurial. The potential of Mercurial is thus lost on them.
Perhaps we should have a brown bag and/or better documentation on getting the most out of Mercurial?
For Gecko/Firefox development, we maintain separate repositories for the trunk and release branches. This introduces all kinds of annoying.
We should not have separate repositories for central, inbound, aurora, beta, release, etc. We should be using some combination of branches and bookmarks and have all the release heads in one repository, just like how the GitHub mirror is configured.
As an experiment, I created a unified Mercurial repository. Each current repository is tracked as a bookmark (there are instructions for reproducing this). Unfortunately, the web interface isn't showing bookmarks (perhaps because the version of Mercurial is too old?), so you'll have to clone the repository to play around. Just run hg bookmarks and e.g. hg up aurora after cloning. Warning: I'm not actively synchronizing this repository, so don't rely on it being up to date.
A Mercurial contributor (who is familiar with Mozilla's development model) suggested we use Mercurial branches for every Gecko release (20, 21, 22, etc). I think this and other uses of branches and bookmarks are ideas worth exploring.
Gecko/Firefox has a complicated code lifecycle and landing process. This could be significantly streamlined if we fully harnessed and embraced the extensibility of Mercurial. While there are some Mozilla-centric extensions (details in my recent post), I don't think they are well known nor used.
I think Mozilla should embrace the functionality of extensions like these (whether they be for Mercurial, Git, or something else) and invest resources in improving the workflows for all developers. Until these tools are obviously superior and advertised, I believe many developers will unknowingly continue to toil without them. And, it's likely hurting our ability to attract and retain new contributors as well.
Mozilla's current usage of Mercurial is far from ideal. It's no wonder people don't like Mercurial (and why some want to switch to Git).
Fortunately, little has to do with shortcomings of Mercurial itself (at least with newer versions). If you want to know why Mercurial isn't working too well for Gecko/Firefox development, most of the problems are self-inflicted or the solutions reside within each of us. Time will tell if we as a community have the will to address these issues.
]]>Two years later and I'm on much better terms with Mercurial. I initially thought it might be Stockholm Syndrome, but after reflection I can point at specific changes and enlightenments that have reshaped my opinions.
I first started using Mercurial in the 1.8 days and thought it was horrible. However, modern releases are much, much better. I've noticed a steady improvement in the quality and speed of Mercurial in the last few years.
If you aren't running 2.5 or later (Mercurial 2.6 was released earlier this month), you should take the time to upgrade today. When you upgrade, you should of course read the changelog and upgrade notes so you can make the most of the new features.
For my workflow, the default configuration of Mercurial out of the box is... far from optimal. There are a number of basic changes that need to be made to satisfy my expectations for a version control tool.
I used to think this was a shortcoming with Mercurial: why not ship a powerful and useful environment out of the box? But, after talking to a Mercurial core contributor, this is mostly by design. Apparently a principle of the Mercurial project is that the CLI tool (hg) should be simple by default and should minimize foot guns. They view actions like rebasing and patch queues as advanced and thus don't have them enabled by default. Seasoned developers may scoff at this. But, I see where Mercurial is coming from. I only need to refer everyone to her first experience with Git as an example of what happens when you don't aim for simplicity. (I've never met a Git user who didn't think it overly complicated at first.)
Anyway, to get the most out of Mercurial, it is essential to configure it to your liking, much like you install plugins or extensions in your code editor.
Every person running Mercurial should go to http://mercurial.selenic.com/wiki/UsingExtensions and take the time to find extensions that will make your life better. You should also run hg help hgrc to view all the configuration options. There is a mountain of productivity wins waiting to be realized.
For reference, my ~/.hgrc. Worth noting are some of the built-in externsions I've enabled:
If I were on Linux, I'd also use the inotify extension, which installs filesystem watchers so operations like hg status are instantaneous.
In addition to the built-in extensions, there are a number of 3rd party extensions that improve my Mozilla workflow:
I'm amazed more developers don't use these simple productivity wins. Could it be that people simply don't realize they are available?
Mozilla has a bug tracking easier configuration of the user's Mercurial environment. My hope is one day people simply run a single command and get a Mozilla-optimized Mercurial environment that just works. Along the same vein, if your extensions are out of date, it prompts you to update them. This is one of the benefits of a unified developer tool like mach: you can put these checks in one place and everyone can reap the benefits easily.
The major differentiator from almost every other version control system (especially Git) is the ease and degree to which Mercurial can be extended and contorted. If you take anything away from this post it should be that Mercurial is a flexible and agile tool.
If you want to change the behavior of a built-in command, you can write an extension that monkeypatches that command. If you want to write a new command, you can of course do that easily. You can have extensions interact with one another - all natively. You can even override the wire protocol to provide new capabilities to extend how peers communicate with one another. You can leverage this to transfer additional metadata or data types. This has nearly infinite potential. If that's not enough, it's possible to create a new branching/development primitive through just an extension alone! If you want to invent Git-style branches with Mercurial, you could do that! It may require client and server support, but it's possible.
Mercurial accomplishes this by being written (mostly) in Python (as opposed to C) and by having a clear API on which extensions can be built. Writing extensions in Python is huge. You can easily drop into the debugger to learn the API and your write-test loop is much smaller.
By contrast, most other version control systems (including Git) require you to parse output of commands (this is the UNIX piping principle). Mercurial supports this too, but the native Python API is so much more powerful. Instead of parsing output, you can just read the raw values from a Python data structure. Yes please.
Since I imagine a lot of people at Mozilla will be reading this, here are some ways Mozilla could leverage the extensibility of Mercurial:
Essentially, Mercurial itself could become the CLI tool code development centers around. Whether that is a good idea is up for debate. But, it can. And that says a lot about the flexibility of Mercurial.
When you consider the previous three points, you arrive at a new one: Mercurial has a ton of future potential. The fact that extensions can evolve vanilla Mercurial into something that resembles Mercurial in name only is a testament to this.
When I sat down with a Mercurial core contributor, they reinforced this. To them, Mercurial is a core library with a limited set of user-facing commands forming the stable API. Since core features (like storage) are internal APIs (not public commands - like Git), this means they aren't bound to backwards compatibility and can refactor internals as needed and evolve over time without breaking the world. That is a terrific luxury.
An example of this future potential is changeset evolution. If you don't know what that is, you should because it's awesome. One of the things they figured out is how to propagate rebasing between clones!
Two years ago I would have said I would never opt to use Mercurial over Git. I cannot say that today.
I do believe Git still has the advantage over Mercurial in a few areas:
However, I believe Mercurial has the upper hand for:
It's worth calling out the major detractors for each.
I think Git's major weakness is its lack of extensibility and inability to evolve (at least currently). Git will need to grow a better extensibility model with better abstractions to compete with Mercurial on new features. Or, the Git community will need to be receptive to experimental features living in the core. All of this will require some major API breakage. Unfortunately, I see little evidence this will occur. I'm unable to find a vision document for the future of Git, a branch with major new features, or interesting threads on the mailing list. I tried to ask in their IRC channel and got crickets.
I think Mercurial's greatest weakness is lack of developer mindshare. Git and GitHub are where it's at. This is huge, especially for projects wanting collaboration.
Of all those points, I want to stress the extensibility and future potential of Mercurial. If hacking your tools to maximize potential and awesomeness is your game, Mercurial wins. End of debate. However, if you don't want to harness these advantages, then I think Git and Mercurial are mostly on equal footing. But given the rate of development in the Mercurial project and relative stagnation of Git (I can't name a major new Git feature in years), I wouldn't be surprised if Mercurial's feature set obviously overtakes Git's in the next year or two. Mind share will of course take longer and will likely depend on what hosting sites like GitHub and Bitbucket do (I wouldn't be surprised if GitHub rebranded as CodeHub or something some day). Time will tell.
I have removed the case study that appeared in the original article because as Mike Hommey observed in the comments, it wasn't a totally accurate comparison. I don't believe the case study significantly added much to the post, so I likely won't write a new one.
From where I started with Mercurial, I never thought I'd say this. But here it goes: I like Mercurial.
I started warming up when it became faster and more robust in recent versions in the last few years. When I learned about its flexibility and the fundamentals of the project and thus its future potential, I became a true fan.
It's easy to not like Mercurial if you are a new user coming from Git and are forced to use a new tool. But, once you take the time to properly configure it and appreciate it for what it is and what it can be, Mercurial is easy to like.
I think Mercurial and Git are both fine version control systems. I would happily use either one for a new project. If the social aspects of development (including encouraging new contributors) were important to me, I would likely select Git and GitHub. But, if I wanted something just for me or I was a large project looking for a system that scales and is flexible or was looking to the future, I'd go with Mercurial.
Mercurial is a rising star in the version control world. It's getting faster and better and enabling others to more easily innovate through powerful extensions. The future is bright for this tool.
]]>The raw data is available.
]]>SQLite.jsm is an abstraction around the low-level Storage APIs that you would have used before. However, it eliminates most of the footguns and makes it easier to write code that doesn't jank the browser and is less prone to memory leaks. It even has an API to free as much memory as possible from the current connection!
If you currently use SQLite via Storage, I highly recommend taking a look at SQLite.jsm - especially if you are using synchronous APIs.
If you are investigating SQLite for the storage needs of your add-on or browser feature, please keep in mind that SQLite can incur lots of filesystem I/O and may run slowly on old machines (especially with magnetic hard drives) and especially with its default configuration. You may be interested in low-level file I/O using OS.File instead.
If you insist on using SQLite, please educate yourself on and then seriously consider using Write-Ahead Logging mode on your database. Some detailed discussion on SQLite behavior as it pertains to Firefox can be found in bug 830492. I hope to eventually incorporate more sane by default connection options into SQLite.jsm to make it easier for add-ons and browser features to have the least-impactful behavior by default (e.g. enable WAL by default). Until then, please, please, please research PRAGMA statements to optimize how your SQLite database runs so it has as little performance overhead as possible. Also consider dropping into an IRC channel on irc.mozilla.org and asking for advice from one of the many who have fallen into SQLite's many performance pitfalls (including me).
]]>I thought this was unacceptable. I wanted to enable people to use their tool of choice (Git) to develop Firefox. So, I did what annoyed engineers do when confronted with an itch: I scratched it.
When I started tackling this problem, I had little knowledge of the problem space other than the problem statement: converting from Mercurial to Git is prohibitively slow and that the slow tool was hg-git. My challenge was thus to make hg-git faster.
When confronted with a performance problem, one of the first things you do is identify the source of the bad performance. Then, you need to acertain whether that is something you have the ability to change.
This often starts by answering some high-level questions then drilling down into more detail as necessary. For a long-running system tool like hg-git, I start with the top test: how much CPU, memory, and I/O is the process utilizing?
In the case of hg-git, we were CPU bound. The Python process was consistently pegging a single CPU core while periodically incurring I/O (but not nearly enough to saturate a magnetic disk). This told me a few things. First, I should look for bottlenecks inside Python. Second, I should investigate whether parallel execution would be possible. The latter is especially important these days because the trend in processors is towards more cores rather than higher clock speeds. It's no longer acceptable to let increases in clock speed or cycle efficiency bail you out: if you want a CPU bound process to run as fast as possible, it's often necessary to involve all available CPU cores.
Once I diagnosed CPU as the limiting factor, I pulled out the next tool in the arsenal: a code profiler. I quickly discovered exactly where the conversion was taking the most CPU time. As feared, it was in the export Mercurial changeset to Git commit function. Specifically, profiling had flagged the conversion of Mercurial manifests to Git trees and blobs. Furthermore, most of the time was spent in functions in Mercurial itself (Mercurial is implemented in Python and hg-git calls into it natively) and Dulwich (a pure Python implementation of Git). So, I was either looking at deficiencies or Mercurial and/or Dulwich, a bad conversion algorithm in hg-git, or both. To know which, I would need a better grasp on the internal storage models of Mercurial and Git.
To understand why conversion from Mercurial to Git was slow, I needed to understand how each stored data internally. My hope was that if attained better understanding I could apply the knowledge to assess the algorithm hg-git was using and optimize it, hopefully introducing parallel execution along the way.
I already had a fairly good understanding of how Git works internally. And, it's quite simple really. The Git Internals chapter of the Pro Git is extremely useful. While I encourage readers to read all of the Git Objects section, the gist is:
When you introduce a new file that hasn't been seen before, a new blob is added to storage. That blob is referenced by a tree. When you update a file, a new tree is created referring to the new blob that was created.
Things get a little complicated when you consider directories. If you update the file foo/bar/baz.c, the tree for foo/bar changes (because the SHA-1 of baz.c changed). And, the SHA-1 for the foo/bar tree changes, so the bar entry in foo's tree changes, changing the SHA-1 for the root tree.
That's essentially how Git addresses commits, directories, and files. If you don't grok this, please, please read the aforementioned page on it - it may even help you better grok Git!
Unlike Git, I didn't really have a clue how Mercurial worked internally. So, I needed to do some self-education here.
The best resource for Mercurial's storage model I've found is the Behind the Scenes chapter from Mercurial: The Definitive Guide. The gist is:
From a high level, Mercurial's storage model is very similar to Git's. They both address files by hashing their content. Where Git uses multiple tree objects to define every file in a commit, Mercurial has a single manifest containing a flat list. Aside from that, the differences are mostly in implementation details. These are important, as we'll soon see.
Armed with knowledge of how Git and Mercurial internally store data, I was ready to analyze how hg-git was performing conversion from Mercurial to Git. Since profiling revealed it was the convert a single changeset into Git commit function that was taking all the time, I started there.
In Python (but not the actual Python), the algorithm was essentially:
def export_changeset_to_git(changeset, git, already_converted): """Receives the Mercurial changeset and a handle on a Git object storre.""" # This is an entity that helps us build Git tree objects from # paths and blobs. The logic is at # https://github.com/jelmer/dulwich/blob/2a8548be3b1fd4a1ae7d0436dce91611112c47c2/dulwich/index.py#L298 tree_builder = TreeBuilder() for file in changeset.manifest: blob_id = already_converted.get(file.id, None) if blob_id is None: blob = Blob(file.data()) git.store(blob.id, blob.content) already_converted[file.id] = blob.id blob_id = blob.id tree_builder.add_file(file.path, blob_id, file.mode) for tree in tree_builder.all_trees(): git.store(tree.id, tree.content) root_tree = tree_builder.root_tree # And proceed to build the Git commit and insert it.
On the face of it, this code doesn't seem too bad. If I were writing the functionality from scratch, I'd likely do something very similar. So, why is it so slow?
As I mentioned earlier, profiling results had identified Mercurial and Dulwich as the hot spots. The Mercurial hotspot was in iteration over the files in the manifest. And the Dulwich offender with Git tree object construction. By why?
First, it turns out that iterating a manifest the way hg-git was isn't exactly performant. I never traced all the gory details, but I'm pretty sure every time it accessed the file context through the change context there was I/O involved. Not good, especially if you may not need the information contained if it was already cached!
Second, it turns out that creating Git tree objects in Dulwich is rather slow. And, the problem is magnified when converting large repositories - like mozilla-central (Firefox's canonical repository).
So, I was faced with a decision: make Mercurial and/or Dulwich faster or change hg-git. Since improving these would have benefits outside of hg-git, I initially went down those roads. However, I eventually abandonded the effort because of effort involved. And, in the case of Dulwich, improving things would likely require rewriting some pieces in C - not something I cared to do nor something that the Dulwich people would likely accept since Dulwich is all about being a pure Python implementation of Git! And in hindsight, this was the right call. Mercurial and Dulwich are fast enough - it's hg-git that was being suboptimal.
I was faced with two problems: don't mass iterate over manifests and don't mass generate Git trees. Both were seemingly impossible to avoid because both are critical to converting a Mercurial changeset to Git.
I thought about this problem for a while. I experimented with numerous micro benchmarks. I engaged the very helpful Mercurial developers on IRC (thanks everyone!). And, I eventually arrived at what I think is an elegant solution.
When I took a step back and looked at the larger problem of exporting Mercurial changesets to Git, I realized it would be beneficial in terms of efficiency for the conversion to be more aware of what had occurred before. Before I came along, hg-git was asking Mercurial for the full state of each changeset for each changeset conversion. When you think about it in low-level operations, this is extremely inefficient. Let's take Git trees as an example.
When you perform a commit, only the trees - and their parents - that had modified files will change. All the other trees will be identical across commits. For large repositories (in terms of files and directories) like mozilla-central, the number of static trees across small commits is quite significant compared to changed trees. The overhead of computing all these trees is not insignificant!
Instead of throwing away all the trees and file context information between changeset exports, what if I preserved it and reused it for the next changeset? I think I'm on to something...
To minimize the work performed when exporting Mercurial changesets to Git, I implemented a standalone class that can emit Mercurial changeset deltas in terms of Git objects. Essentially, it caches a Git tree representation of a Mercurial manifest. When you feed a new Mercurial changeset into it, it asks Mercurial to compare those changesets using the same API used by hg status. This API is efficient and returns the information I care about: the paths that changed. Once we have the changed files, we simply reflect those changes in terms of updating Git trees.
If a file changes or is added, we emit a blob. If a tree changes, we emit the new tree object. When the consumer has finished writing the set of new objects to Git, it asks for the SHA-1 of the root tree. (Up until this point the consumer is not aware of what any of the emitted objects actually are - just that they likely need to be added to storage.) It then uses the SHA-1 of the root tree to construct the commit. Then it moves on to the next changeset.
The impact of this change is significant. On my computer, converting Mercurial's own Mercurial repository Git went from 21:07 to 8:14 on my i7-2600k. mozilla-central is even more drastic. The first 200 commits (the first commit was a large dump from CVS) took 8:17 before and now take 2:32. I don't have exact numbers from newer commits, but I do know they were at least twice as slow as the initial commits and showed an even more drastic speedup.
But I was just getting started.
The initial implementation wasn't very efficient in terms of reducing tree object calculations. I changed that earlier today when I submitted a patch for consideration that only calculates tree changes for trees that actually changed. I also removed some needless sorting on the order of export operations. This second patch reduced conversion of Mercurial's repository down to 5:33. Even more impressive is that mozilla-central's changesets are now exporting almost 4x faster with this patch alone. The first 200 changesets now export in 42s (down from 2:32 which is down from 8:17). This is mostly due to the overhead of reprocessing non-dirty trees on every export.
And I'm not through.
As part of building the standalone incremental changeset exporter, one of the goals in the back of my mind was to eventually have things execute in parallel.
In my personal development branch I have a patch to perform Mercurial changeset export on multiple cores. Essentially hg-git fires up a bunch of worker processes and asks each to export a consecutive range of changesets. The workers writes new Git objects into Git and then tells the coordinator process the root tree SHA-1 corresponding to each Mercurial changeset. The coordinator process then uses these root tree SHA-1's to derive Git commit objects (you can't create the commit object until you know the SHA-1 of the commit's parents).
The blob and tree exporting on separate processes makes Mercurial to Git export scale out to however many cores you feel like throwing at it. When 32 core machines come around, you can convert using all available cores and the speedup should roughly be linear with the number of cores.
I'm still working out some kinks in the multiple processes patch (the multiprocessing module is very difficult to get working on all platforms and I don't want to break hg-git when it lands). But, Ehsan Akhgari has been using it to power the GitHub mirror of mozilla-central for months without issue. (His use of these patches freed up the CPU required to support conversion of more project branches on the Git mirror. And, he's still not using the 4x improvement patch I wrote today - he will shortly - so who knows what improvements will stem from that.)
With all the patches applied, hg-git now feels like a Ferrari when exporting Mercurial changesets to Git. Conversion of Mercurial's repository now takes 1:25 (down from 21:07). Conversion of mozilla-central has gone from 6+ days to about 3 hours! More importantly, ongoing conversions feel somewhat snappy now.
With the patch today, I'd say optimization of exporting Mercurial changesets is nearing its limits. There are a few things I could try that may net another 2 or 3x improvement. But, I think the ~50x improvement I've already attained (at least for mozilla-central) is pretty damn good and good enough for most users. (Part of performance optimization is knowing when is good enough and stopping before you invest excessive time in the long tail.)
There is one giant refactor that could likely net a significant win for Git export. However, it requires optimizing for initial export over recurring incremental export (which is why I have little interest in it). Incremental export incurs a lot of random I/O accessing Mercurial filelogs and extracting specific file revisions as they are needed. An optimal export would iterate over the filelogs and export Git blobs from each filelog in the sequence they occur in within the filelogs. It would cache the file node to blob SHA-1. After all blobs are exported, the mappings would be combined and distributed to all workers. Then, tree export would occur in parallel largely under the existing model modulo blob writing. This would minimize overall I/O and work in Mercurial and would likely be significantly faster. However, it's mostly useful for initial export and IMO not worth implementing. (It's possible to employ a variation for incremental export that iterates over filelogs and exports not-yet-seen revisions. Perhaps I will investigate this some day.)
Now that I've tackled Mercurial to Git conversion, it's very tempting to work magic on the inverse: converting Git commits to Mercurial changesets. While I haven't looked at this problem in detail, I already know it will be at least slightly more challenging.
The reason is parallelization. With Mercurial export, I have each child process reading directly from Mercurial and writing directly to Git. There are no locks involved. There is just a coordinator that ensures minimum redundant work among workers. There is some redundant work, sure. But, the alternative would be lots of locking and/or exchange of state across processes - not cheap operations! Furthermore, the writes into Git can occur in any order (since Git is just a key-value store). The only hard requirement is a child commit must come after its parent (because you need the parent commit's SHA-1). And, single-threaded insert of commit objects isn't a big deal because you can crank through hundreds of them per second (it might even be over 1000/s on my machine).
Mercurial's storage implementation does not afford me the same carelessness with regards to writing into storage. Since Mercurial uses shared files for individual file and manifest history, we have a contention problem. We could lock files when writing to them. However, these files (revlogs in Mercurial speak) also use transparent delta compression. You get the best performance/compression when changes are written in the order they actually occured in (at least in the typical case).
To optimally write to Mercurial you need to order inserts. This means parallel reads from Git (in separate worker processes) would be very difficult to implement. Doable, sure, but you're looking at a lot of transferred state and ordering. This likely involves a lot more memory and CPU usage.
The best idea I've come up with so far is a single process that reads off Git commits and iterates trees. It hashes the paths of seen files to a consistent worker process which then pulls the blob from Git's storage and inserts it into the filelog. You don't need to lock filelogs because only one worker owns a specific path. Workers report the blob's corresponding file node to another process which then assembles manifests, writes manifests in order, and finally creates and writes changesets. Unfortunately, the worker processes are just doing blob I/O. There is no parallel processing of Git tree calculation or Mercurial manifests. Given this was a significant source of slowness exporting to Git, I worry the inverse will be true. Although, the problem with Git was tree creation and it was due to the volume. Since there is only 1 manifest per changeset, perhaps it won't be as bad.
While I've brainstormed a solution, I have no concrete plans to work on Git to Mercurial conversion. The impetus for me working on Mercurial to Git speedups was that I and a number of other Mozilla people were personally impacted. If the same is true for Git to Mercurial slowness, I could invest a few hours the next time I'm sick and bored over the weekend.
Converting Mercurial repositories to Git with hg-git is now significantly faster. If you thought it was too slow before, grab the latest code (from either the official repository or my personal branch) and enjoy.
]]>Most of the data from Mozilla's continuous integration is available on public servers, notably ftp.mozilla.org. This includes compiled binaries, logs, etc.
While there are tools that can sift through this mountain of data (like TBPL), they don't allow ad-hoc queries over the raw data. Furthermore, these tools are very function-specific and there are many data views they don't expose. This missing data has always bothered me because, well, there are cool and useful things I'd like to do with this data.
This itch has been bothering me for well over a year. The persistent burning sensation coupled with rain over the weekend caused me to scratch it.
The product of my weekend labor is a system facilitating bulk storage and analysis of Mozilla's build data. While it's currently very alpha, it's already showing promise for more throrough data analysis.
Essentially, the tool works by collecting the dumps of all the jobs executed on Mozilla's infrastructure. It can optionally supplement this with the raw logs from those jobs. Then, it combs through this data, extracts useful bits, and stores them. Once the initial fetching has completed, you simply need to re-"parse" the data set into useful data. And, since all data is stored locally, the performance of this is not bound by Internet bandwidth. In practice, this means that you can obtain a new metric faster than would have been required before. The downside is you will likely be storing gigabytes of raw data locally. But, disks are cheap. And, you have control over what gets pulled in, so you can limit it to what you need.
Please note the project is very alpha and is only currently serving my personal interests. However, I know there is talk about TBPL2 and what I have built could evolve into the data store for the next generation TBPL tool. Also, most of the work so far has centered on data import. There is tons of analysis code waiting to be written.
If you are interested in improving the tool, please file a GitHub pull request.
I hope to soon blog about useful information I've obtained through this tool.
]]>Essentially, Matt changed the mach driver (the script in the root directory of mozilla-central) so instead of having it look in hard-coded relative paths for all its code, it walks up the directory tree and looks for signs of the source tree or the object directory.
What this all means is that if you have the mach script installed in your $PATH and you just type mach in your shell from within any source directory or object directory, mach should just work. So, no more typing ./mach: just copy mach to ~/bin, /usr/local/bin or some other directory on your $PATH and you should just be able to type mach.
Unfortunately, there are bound to be bugs here. Since mach traditionally was only executed with the current working directory as the top source directory, some commands are not prepared to handle a variable current working directory. Some commands will likely get confused when it comes resolving relative paths, etc. If you find an issue, please report it! A temporary workaround is to just invoke mach from the top source directory like you've always been doing.
If you enjoy the feature, thank Matt: this was completely his idea and he saw it through from conception to implementation.
]]>