A new API to graph server was just enabled today; it lets you fetch performance data for a given revision. E.g. http://graphs.mozilla.org/api/test/runs/revisions?revision=10d2046d2b64&revision=d16525937c8b Documentation is here, and the gory implementation details are in Bug 551732. Enjoy!
Yesterday was a bit of an overwhelming day. After getting home at 1am after a long bus ride home, I was unwinding by catching up on some news and email. I came across these two links, both of which really lifted my mood. The first, Grokking the Zen of the Vi Wu-Wei, talks about a programmer's journey from emacs to BBEdit to vim. This post is a great read in and of itself, but what's really worth it, is the link around the middle of the post to http://stackoverflow.com/questions/1218390/what-is-your-most-productive-shortcut-with-vim/1220118#1220118. This was truly a joy to read. Definitely the best answer I've ever seen on Stack Overflow, and quite possibly the best discussion of vi I've ever read. It taught me a lot, but I enjoyed reading it for more than that. It was almost like being on a little adventure, discovering all these little hidden secrets about the neighbourhood you've been living in for years. Like I said, it was 1am. The second, The Pope, the judge, the paedophile priest and The New York Times, gave me some reassurance that things aren't always as they seem as reported by the media. Regardless of how you feel about the Church or the Pope, it seems that journalistic integrity has fallen by the wayside here. From the article:
Fr Thomas Brundage, the former Archdiocese of Milwaukee Judicial Vicar who presided over the canonical criminal case of the Wisconsin child abuser Fr Lawrence Murphy, has broken his silence to give a devastating account of the scandal – and of the behaviour of The New York Times, which resurrected the story. It looks as if the media were in such a hurry to to blame the Pope for this wretched business that not one news organisation contacted Fr Brundage. As a result, crucial details were unreported.The entire article is worth a read.
It seems like it was ages ago when I posted about profiling buildbot.
One of the hot spots identified there was the dataReceived call. This has been sped up a little bit in recent versions of twisted, but our buildbot masters were still severely overloaded.
It turns out that the buildbot slaves make a lot of RPC calls when sending log data, which results in tens of thousands of dataReceived calls. Multiply that by several dozen build slaves sending log data peaking at a combined throughput of 10 megabits/s and you've got an awful lot of data to handle.
By adding a small slave-side buffer, the number of RPC calls to send log data is drastically reduced by an order of magnitude in some tests, resulting in a much better load situation on the master. This is good for us, because it means the masters are much more responsive, and it's good for everybody else because it means we have fewer failures and wasted time due to the master being too busy to handle everything. It also means we can throw more build slaves onto the masters!
The new code was deployed towards the end of the day on the 26th, or the end of the 12th week.
I'm happy to announce that we've started publishing some build data we've been collecting for the past several months. If you're interested in build data, like what jobs get run on which machines, and how long they take, then the files at http://build.mozilla.org/builds will be of interest to you. These are JSON dumps of build data collected for most of our systems since October. There is one file per 24-hour period, as well as one file for the past 4 hours of data. If you want to look at the format, the 4 hour file is easier to read; the other files don't use any extra white space. All times should be interpreted as unix timestamps (seconds since Jan 1, 1970 00:00:00 UTC). The "requesttime" is a best-effort calculation of when a build was requested based on when the revision in question was pushed to hg, or when the unit test or talos run was triggered. That's it! This should be enough data to get started writing some dashboards and other visualizations. I'd love to hear how people are using this data, and if there's anything missing from the data provided that would be useful.
One metric Release Engineering focuses on a lot is the time between a commit to the source tree and when all the builds and tests for that revision are completed. We call this our end-to-end time. Our approach to improving this time has been to identify the longest bits in the total end-to-end time, and to make them faster.
One big chunk of the end-to-end time used to be wait times: how long a build or test waited for a machine to become free. We've addressed this by adding more build and test slaves into the pool.
We've also focused on parallelizing the entire process, so instead of doing this:
we now do this: 
(not exactly to scale, but you get the idea)
(the break in the graph is the time when mozilla-central was closed due to the PGO link failure bug)
We started rolling the new builders into production on February 22nd (where the orange diamonds start on the graph). You can see the end-to-end time immediately start to drop. From February 1st to the 22nd, our average end-to-end time for win32 on mozilla-central was 4h09. Since the 22nd the average time has dropped down to 3h02. That's over an hour better on average, a 26% improvement.
Faster builds means tests can start faster, which means tests can be done sooner, which means a better end-to-end time. It also means that build machines become free to do another build sooner, and so we're hoping that these faster build machines will also improve our wait time situation (but see below).
Currently we're limited to running the build on these machines with -j1 because of random hangs in the build when using -j2 or higher (bug 524149). Once we fix that, or move to pymake, we should see even better improvements.
Simply selecting the right machine for the job reduced our end-to-end time from 3h12 to 2h13, again almost an hour improvement, or 30% better.
Johnathan posted links to 3 scripts he finds useful. His sattap script looked handy, so I hacked it up for linux. Run it to do a screen capture, and upload the image to a website you have ssh access into. The link is printed out, and put into the clipboard. Hope you find this useful!
#!/bin/sh
# sattap - Send a thing to a place
set -e
SCP_USER='catlee'
SCP_HOST='people.mozilla.org'
SCP_PATH='~/public_html/sattap/'
HTTP_URL="http://people.mozilla.org/~catlee/sattap/"
FILENAME=`date | md5sum | head -c 8`.png
FILEPATH=/tmp/$FILENAME
echo Capturing...
import $FILEPATH
echo Copying to $SCP_HOST
scp $FILEPATH ${SCP_USER}@${SCP_HOST}:$SCP_PATH
echo Deleting local copy
rm $FILEPATH
echo $HTTP_URL$FILENAME | xclip -selection clipboard
echo Your file should be at $HTTP_URL$FILENAME, which is also in your paste buffer
As of November 1st, when you push a change to mozilla-central, the following builds and tests get triggered:
|
|
|
|
Continuing our RelEng Blogging Blitz, I'm going to be discussing how and when tests get triggered in our build automation systems.
We've got two basic classes of tests right now: unit tests, and performance tests, a.k.a. Talos. The unit tests are run on the same pool of machines that the builds are done on, while the performance tests are run on a separate pool of around 100 Mac Minis. Both kinds of tests are triggered in similar ways.
For refcounting ("unittest") builds, once the compile step is complete, the binaries are packaged up with make package, the tests are packaged up with make package-tests, the symbols are packaged up with make buildsymbols, and then the whole lot is uploaded to stage.mozilla.org using make upload. Once they're uploaded, we have valid URLs that refer to the builds, tests, and symbols. We then trigger the relevant unit test runs on that build. When a slave is assigned this test run, it then downloads the build, tests, and symbols from stage and starts running the tests.
On mozilla-central, we've also recently started to run unittests on optimized and debug builds. We're hoping to bring this functionality to mozilla-1.9.2 once all the kinks are worked out.
For regular optimized builds, in addition to unittests, we also trigger performance tests on the freshly minted build. OSX builds are currently tested on Tiger and Leopard for mozilla-1.9.1 and mozilla-1.9.2, and on Leopard only for mozilla-central and project branches. Windows builds are tested on XP and Vista, and Linux builds are tested on Ubuntu.
In addition to having tests triggered automatically by builds, the Release Engineering Sheriff can re-run unittests or performance tests on request!
As part of our RelEng Blogging Blitz, I'll give a quick overview of when and how builds get triggered on our main build infrastructure. There are three ways builds can be triggered. The first, and most common way, is when a developer pushes his or her changes to hg.mozilla.org. Our systems check for new changes every minute or so, and put new changes into a queue. Once the tree has been quiet for 3 minutes (i.e. no changes for 3 minutes), a new build request is triggered with all queued changes. If there is a free slave available, then a new build starts immediately, otherwise the build request is put in a queue. The second way builds are triggered is via a nightly scheduler. We start triggering builds on branches at 3:02am pacific local time (some branches are triggered at 3:32am or 4:02 am). We run at 3:02am to avoid problems with daylight savings <-> standard time transitions. In the fall there are two 2:59am's when we go back to standard time, and in the spring transition there is no 2:59am. The start times are staggered to avoid slamming hg.mozilla.org, or other shared resources. The last way builds can be triggered is manually. The Release Engineering Sheriff can trigger builds on specific revisions, or rebuild past builds pretty easily, so if you need a build triggered, contact your friendly neighbourhood RelEng Sheriff!
Folks from RelEng are going to be blogging about various bits of our build, test, and release automation infrastructure this week. Stay tuned for more!