Merged
Conversation
This will allow blinkpy to pass its file manager to typ's Artifacts class. Then blinkpy's mock file manager will be able to keep track of artifacts created and prevent writing actual files to the disc. TBR=dpranke@chromium.org Bug: chromium:982208 Change-Id: I728f9aa32e7770fa0fb8e4fd7a0fde11d9434670 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1842408 Commit-Queue: Rakib Hasan <rmhasan@google.com> Reviewed-by: Rakib Hasan <rmhasan@google.com>
This reverts commit a120c4f. Reason for revert: This CL removes the verbose option. However the option is still used in the telemetry_perf_unittests step. This CL is blocking the catapult autoroll. Original change's description: > [Telemetry] Remove --verbose option and Popen hook > > The --verbose option is now handled by Results Processor. The hook was > introduced for debugging of one specific bug and is not necessary anymore. > > Bug: chromium:985712 > Change-Id: I070cffa685890e9610f9b341248739eba201b082 > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1840372 > Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> > Commit-Queue: Mikhail Khokhlov <khokhlov@google.com> TBR=perezju@chromium.org,khokhlov@google.com # Not skipping CQ checks because original CL landed > 1 day ago. NOTRY=True Bug: chromium:985712 Change-Id: I90d3be59146ec54763309ce3566e09a3d6441242 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1843912 Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Rakib Hasan <rmhasan@google.com>
When establishing a DevTools connection, the TracingBackend is also created. Therefore, if Chrome startup tracing has already be enabled, we can take this chance to send a new Tracing.start command to confirm and update settings on the tracing session; rather than waiting to the end close to the StopTracing command. Gladly this also makes the code a bit more readable and easy to understand. Bug: chromium:914092 Change-Id: I59072e437c989375518d94de0e1a5f92f450480c Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1840371 Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Reviewed-by: Eric Seckler <eseckler@chromium.org>
This reverts commit 8305659. Reason for revert: Seems like windows and mac are having trouble too so too early to remove m72. Original change's description: > Use chromium build on linux for dev server tests > > run_dev_server_tests now automatically switches to using chromium builds > when (1) platform is linux and (2) --no-use-local-chrome is specified. > > The update_chrome_reference_binaries scripts is also updated to > automatically upload chromium builds to binary_dependencies. Chromium > builds are fetched from the chromium-browser-snapshots bucket that is > used by bisect_builds.py. Often the exact commit position corresponding > to the channel is not available, so the closest one is chosen. > > I ran the updated script (only for chromium builds by > using --no-update-chrome) because I did not want to include chrome > updates in this patch) and added the changes to chrome_binaries.json. > > All references to m72 channel has been removed. > > Bug: chromium:973847 > Change-Id: Idfa55c06ed6ad469e3ec7de38825287aedcb4f0e > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1787900 > Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> > Reviewed-by: Caleb Rouleau <crouleau@chromium.org> > Reviewed-by: Deepanjan Roy <dproy@chromium.org> > Commit-Queue: Deepanjan Roy <dproy@chromium.org> TBR=perezju@chromium.org,crouleau@chromium.org,dproy@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: chromium:973847 Change-Id: Iab1833d82cdac524ffb141486b026a4a9f464c7b Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1844313 Reviewed-by: Deepanjan Roy <dproy@chromium.org> Commit-Queue: Deepanjan Roy <dproy@chromium.org>
Refactor both AbridgeableStorySetTest and BenchmarkJsonResultsTest tests into RunBenchmarkTest; reusing and adapting FakeBenchmark as needed to avoid code duplication. testArtifactLogsHaveProperPathWithSymlinkedTmp is removed entirely. Is irrelevant for story_runner as it does not manipulate output_dir in any way. (And results_processor does resolve away any symlinks in the path.) This also removes all testing on output_formats, replacing them with checks on intermediate format instead. Bug: chromium:1008852 Change-Id: I2668d5be00f6bcfc4248244b6ceafeafd2a450f6 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1833687 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
This will allow results processor to find them and include them in the processed results. Also allow to add a "description" of each measurement. Press benchmarks do fill this in, which is supported both in legacy values and histograms; so it's probably best not to lose this feature. Bug: chromium:999484 Change-Id: Ibc0bff23b7ff4949d56d6b191b6d20c9258fa16d Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1840835 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
This change introduces an alternative calculation when doing change point detection by avoiding computing the means of the values from the individual attempts. The alternative implementation is there so that we can first count the number of times we encounter cases where the alternative computation has a different finding from the current approach which uses the mean of the results of an attempt. When this change is in production we can set up a temporary log metric to catch the number of times the log message appears, and make a determination on how often the approaches differ -- and whether it's qualitatively better to use the alternative approach. R=abennetts@google.com, simonhatch@chromium.org Bug: chromium:1011273 Change-Id: I0a8b8f485daa863303809ef243f90d8491efd90e Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1839339 Commit-Queue: Dean Berris <dberris@chromium.org> Auto-Submit: Dean Berris <dberris@chromium.org> Reviewed-by: Simon Hatch <simonhatch@chromium.org>
This adds more UI to the job creation dialog to choose the measurement to bisect on, using the results of the describe endpoint. This also reworks the UI layout a bit, makes the whole thing 2 columns. BUG=chromium:1005961 Change-Id: I1b34f5d2ec6dc7a55ac1300613ab904103ea5388 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1845472 Commit-Queue: Dean Berris <dberris@chromium.org> Reviewed-by: Dean Berris <dberris@chromium.org>
Make the frozen job threshold more aggressive. We used to wait 6 hours of no activity before considering a job is frozen, so now we're reducing that to 1h for the following reasons: - Given that we have a scheduler that's able to enforce exclusivity of the pool now, this means we are potentially making pools idle for 6 hours before we consider any action. - Retrying jobs are a good idea anyway, if we fail to find/make progress for a number of times. Calling a job "failed" sooner is better than potentially waiting too long before letting other jobs run sooner. R=abennetts@google.com, simonhatch@chromium.org Bug: chromium:1011728 Change-Id: I3d8b11b9c66a2ff941635dc5e38b6724c91d516c Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846580 Auto-Submit: Dean Berris <dberris@chromium.org> Reviewed-by: Andrew Bennetts <abennetts@google.com> Commit-Queue: Andrew Bennetts <abennetts@google.com>
Make the implementation of the _LogAlternativeComparison function more readable. R=abennetts@google.com, simonhatch@chromium.org Bug: chromium:1011273 Change-Id: I14db2cb33bb13f05db741ae98a68ff3c86f1a502 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846579 Auto-Submit: Dean Berris <dberris@chromium.org> Reviewed-by: Andrew Bennetts <abennetts@google.com> Commit-Queue: Andrew Bennetts <abennetts@google.com>
These are no longer in use after crrev.com/702804. Bug: chromium:921000 Change-Id: Ibd2609dae6fa45ba09a4878197cd25c65e6214e5 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1840373 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
The --verbose option is now handled by Results Processor, so to avoid duplication it's moved from browser_options to run_tests. Tests run from Catapult CQ now pass -v option in the same way as telemetry_unittests in Chromium do. Also the Popen hook that was introduced for debugging one specific bug is removed because it's not needed anymore. Bug: chromium:985712 Change-Id: Ib335ced67ae4f586a8bd93525cfe72512e886180 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1845211 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Mikhail Khokhlov <khokhlov@google.com>
It turns out results2.py depended on functions meant to be internal-only in the read_value quest. Since we've refactored those names to not have the leading underscore, we started failing in production for getting the results of jobs attempting to generate the new version of the results viewer. Also, it turns out, we didn't catch this because there are no unit tests covering this part of the results2 implementation. R=abennetts@google.com, simonhatch@chromium.org TBR=abennetts@google.com,simonhatch@chromium.org Bug: chromium:1012187 Change-Id: Ie5e33d6197f410f95eac3cb594253a2192a21b5e Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846585 Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Dean Berris <dberris@chromium.org> Auto-Submit: Dean Berris <dberris@chromium.org>
This change makes the commit cache only use memcache and the in-memory cache instead of writing through to Datastore. For the purpose of a cache, memcache and the in-memory cache should fine without needing to ensure transactional semantics for the operations on the commit cache. This change also allows us to reduce the amounts of calls to the gitiles service when resolving commits by checking the cache first if we've seen these changes recently. We also resolve a deadlock in the AsDict() member of the Commit object which relies on the computed `result_url` property, but also calls into an independent transaction. We do this by caching the computed repository url once when we've created it in the object. We can apply this transformation for the other cache-only objects we have in Datastore. R=abennetts@google.com, simonhatch@chromium.org TBR=simonhatch@chromium.org Bug: chromium:1006888 Change-Id: I92864263a6ed2b2a14107f410ed0d640444c030a Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846584 Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Dean Berris <dberris@chromium.org> Auto-Submit: Dean Berris <dberris@chromium.org>
More changes in results2 because we don't have test coverage. R=abennetts@google.com, simonhatch@chromium.org TBR=simonhatch@chromium.org Bug: chromium:1012187 Change-Id: I0ba8a50b286bebb70234b51f9c8eb861fbb8f276 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846587 Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Dean Berris <dberris@chromium.org> Auto-Submit: Dean Berris <dberris@chromium.org>
This change is a stop-gap change to allow clients to use OAuth instead of IP whitelisting for the /add_point API. This lets users who are still using the API to migrate to OAuth first before switching to the HistogramSet based APIs. R=abennetts@google.com, simonhatch@chromium.org Bug: v8:9422 Change-Id: I6d5f38143d4dfbeed9fb5f09e956d445cc4269ab Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846576 Reviewed-by: Simon Hatch <simonhatch@chromium.org> Commit-Queue: Simon Hatch <simonhatch@chromium.org> Auto-Submit: Dean Berris <dberris@chromium.org>
If an output_dir is provided, also set the intermediate_dir to a reasonable default. All clients wanting to read the intermediate results will need this, so let's provide it by default for all. Bug: chromium:1012227 Change-Id: Id8fa06634be929817579f14ba4ca4b60abb57dd7 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1847392 Commit-Queue: Caleb Rouleau <crouleau@chromium.org> Auto-Submit: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
This CL will create an interface for which the Artifacts class can interact with the files on disc. The API of this file manager will follow closely the API of blinkpy's file manager. This CL also adds the optional force_overwrite parameter to CreateArtifact, so that it can force the function to skip checking if the file exists. Bug: chromium:982208 Change-Id: I3ed666bcb7edb029d8637e042c9fa2e286766160 TBR=dpranke@chromium.org R=isamsonov@google.com Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1846732 Commit-Queue: Rakib Hasan <rmhasan@google.com> Reviewed-by: Rakib Hasan <rmhasan@google.com>
1. Instead of duplicating the CanRunOnPlatform code, use it directly from commands.py. 2. "run_benchmark list" should figure out which benchmarks are supported, and it should not depend on whether stories happen to be temporarily disabled or not. This change allows us to delete some code for expectations from this file. This lays the groundwork for eventually moving expectations argument support into story_filter.py so it is all in the same place. Then fix the tests. The only way to disable benchmarks from the perspective of run_benchmark list is for them to list SUPPORTED_PLATFORMS now, so we can replace the expectations file code with this, and it is easier and simpler. Note that it would be great to make these unit tests not be change detector tests, but that is beyond the scope of this changelist. Bug: chromium:985103 Change-Id: I5bc085e47e071b04f7e22bfd85ccbf303b413513 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1779351 Commit-Queue: Caleb Rouleau <crouleau@chromium.org> Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org>
We apparently trip the gitiles service throttling when asking for this too often. It's not that important to get absolute ToT, so we can just cache the result of this with a reasonably short expiration. Bug: chromium:1012200 Change-Id: I447ba837db42103562d3f484af263b5f1205a3d8 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1848592 Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Dean Berris <dberris@chromium.org>
The former component no longer exists. Change-Id: If0e05ce6f055700467605a26c3ea2c78e3dc86e0 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1847393 Auto-Submit: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Bug: chromium:1011813 Change-Id: Ic6045f234b6a2471ee8f463c69361c8866f232bf Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1847391 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Mikhail Khokhlov <khokhlov@google.com>
update_chrome_reference_binaries script no longer crashes if there is a missing build. It prints a warning and updates the rest of builds instead. See bug more details. Bug: chromium:1012741 Change-Id: I0aa61bbc488d7b1592207f4ad1a517def2b962c0 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1849164 Commit-Queue: Deepanjan Roy <dproy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
run_dev_server_tests now uses chromium when --no-use-local-chrome is specified. The update_chrome_reference_binaries scripts is also updated to automatically upload chromium builds to binary_dependencies. Chromium builds are fetched from the chromium-browser-snapshots bucket that is used by bisect_builds.py. Often the exact commit position corresponding to the channel is not available, so the closest one is chosen. Included changes to chrome_binaries.json after updating chromium builds. References to m72 has been removed. This CL builds on top of https://crrev.com/c/1787900, which was reverted. Bug: chromium:973847 Change-Id: I3a54f59930ba335e4e2f1baa5db877cc147f9d1f Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1845657 Commit-Queue: Deepanjan Roy <dproy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
Easier to do code review with syntax highlighting. R=crouleau@chromium.org Change-Id: I6d6bc513ccd4c7367c43a7ad4fed780ad504736d Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1854007 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Deepanjan Roy <dproy@chromium.org>
This CL makes changes to typ so that the Artifacts class can use a Host or FakeHost instance to interact with files. It also adds a validation check for the AddArtifacts method to make sure duplicates are not added to an artifact's list. Bug: chromium:982208 Change-Id: I19a28d5ec98c9711a74e21d5944ca417f7517403 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1848395 Commit-Queue: Rakib Hasan <rmhasan@google.com> Reviewed-by: Dirk Pranke <dpranke@chromium.org>
This catapult-side change enables the Chromium side change https://chromium-review.googlesource.com/c/chromium/src/+/1854794 . That change enables the final Telemetry side change https://chromium-review.googlesource.com/c/catapult/+/1839176 The environment argument will later be used to pass project config information to story_runner, which will pass it to story_filter, which will find the expectations file path in it and use that to decide which stories to skip. Bug: chromium:985103 Change-Id: Icecb8ca34427e7f46e7a66d3988b038f20cc8042 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1854882 Auto-Submit: Caleb Rouleau <crouleau@chromium.org> Reviewed-by: Rakib Hasan <rmhasan@google.com> Commit-Queue: Rakib Hasan <rmhasan@google.com>
This is to remove inconsistency between legacy_unit_info.py and legacy_unit_info.html. HTML version, used in chart_json_converter, has 'ms' defined as legacy unit, while python one doesn't. Bug: chromium:981349 Change-Id: I9ce13c55bdcb97e9b79301bbd411b32a4cf38e25 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1853224 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Mikhail Khokhlov <khokhlov@google.com>
We need them in Results Processor to produce correct diagnostics for ad-hoc measurements. Bug: chromium:981349 Change-Id: Ic68671d801ad9d097655ace9bdfa96c3aaa2e83a Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1851711 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Mikhail Khokhlov <khokhlov@google.com>
This reverts commit 8330c31. Reason for revert: Identified as root cause of go/chops-pm-155. Original change's description: > Dashboard - Force test metadata to update sheriff on upload > > There are instances where this goes out of sync, but the entire system > will be removed and replaced with the sheriff redux work. Until then, > we'll manually force an update here to at least ensure that we stop > alerting on ghost paths. > > Bug: chromium:1006252 > Change-Id: I7b104638015f05142c7457c0e6bb1d19a18ffc23 > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1818786 > Commit-Queue: Dean Berris <dberris@chromium.org> > Reviewed-by: Dean Berris <dberris@chromium.org> > Auto-Submit: Simon Hatch <simonhatch@chromium.org> TBR=simonhatch@chromium.org,dberris@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: chromium:1006252, chromium:1013984 Change-Id: I7656a8f5c37311ea391b359982740dcd2b945862 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1859297 Reviewed-by: Andrew Bennetts <abennetts@google.com> Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Andrew Bennetts <abennetts@google.com>
Add a button on the top toolbar to allow filtering flow events based on category. Fixed: chromium:1018750 Change-Id: I5de3943ef1d1ada749c0241299f6e0f8acd60a9e Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1879613 Commit-Queue: Saman Sami <samans@chromium.org> Reviewed-by: Deepanjan Roy <dproy@chromium.org> Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Deduplicates the minidump symbolization code for desktop and ChromeOS by moving it into a separate set of *MinidumpSymbolizer classes. Bug: chromium:994274 Change-Id: I50cf0c0ff31f2f8934f51c2d342f73e9b405663a Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1884465 Commit-Queue: Brian Sheedy <bsheedy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
The concept of "tir_label" (a timeline interaction record label) no longer exists in Telemetry. It was replaced with a simpler "grouping_label" which just allows grouping stories into meaningful groups. This CL updates pinpoint code to use the new name, and be ready for clients that start using the new name in requests. Bug: chromium:974237 Change-Id: Ic741e0e2a982ea20aa3e6b11657dcb6f06970526 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1873701 Commit-Queue: Dean Berris <dberris@chromium.org> Reviewed-by: Dean Berris <dberris@chromium.org>
The binary was created by using configuration in update_v8: ``` gn gen out/Default \ --args="is_debug=false icu_use_data_file=false v8_use_external_startup_data=false v8_static_library=true" ``` R=ulan@chromium.org,dproy@chromium.org CC=crouleau@chromium.org,perezju@chromium.org Bug: chromium:1001953 Change-Id: Iba65d3e60603a09e47a7f23733bac84733303a63 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1878748 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Tamer Tas <tmrts@chromium.org> Auto-Submit: Tamer Tas <tmrts@chromium.org>
This is to be replaced by a --max-values-per-test-case flag in Results Processor. Needs this CL to land first: https://chromium-review.googlesource.com/c/chromium/src/+/1879207 Bug: chromium:1001038 Change-Id: Iea27f0000e25765139667a8b99ea5eca934ed01c Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1879209 Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
The are no longer used. Only Results Processor creates histograms now. TBR=crouleau@chromium.org Bug: chromium:973837 Change-Id: Ieed6b186749f302d2573a2d76e47a5eb4ab1c102 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1879216 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Rename isExpected -> expected and add resultId. Bug: chromium:1018248 Change-Id: I3d159bc32999e45e60397810ec51c7ff4a4b3114 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1887030 Commit-Queue: Mikhail Khokhlov <khokhlov@google.com> Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Nodir Turakulov <nodir@chromium.org>
This reverts commit 090f35f. Reason for revert: Breaks Windows https://crbug.com/1019286 Original change's description: > Deduplicate desktop/CrOS minidump code > > Deduplicates the minidump symbolization code for desktop and ChromeOS by > moving it into a separate set of *MinidumpSymbolizer classes. > > Bug: chromium:994274 > Change-Id: I50cf0c0ff31f2f8934f51c2d342f73e9b405663a > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1884465 > Commit-Queue: Brian Sheedy <bsheedy@chromium.org> > Reviewed-by: Caleb Rouleau <crouleau@chromium.org> TBR=crouleau@chromium.org,bsheedy@chromium.org Change-Id: I7980f5d7f63d8be137c43b1506c63d1037776f30 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: chromium:994274 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1887712 Reviewed-by: Brian Sheedy <bsheedy@chromium.org> Commit-Queue: Brian Sheedy <bsheedy@chromium.org>
Chromium mac builds do not seem to have keystone bundled with them. Previously the update script was crashing trying to do that. Change-Id: I0c883f58deea5c1fa08e0e86e8b19ca0883a41df Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1888058 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Deepanjan Roy <dproy@chromium.org>
This is a reland of 090f35f Changes: 1. Fix _GetCdbPath not being defined on Windows. 2. Moved Windows-specific symbolization code into DesktopMinidumpSymbolizer. Original change's description: > Deduplicate desktop/CrOS minidump code > > Deduplicates the minidump symbolization code for desktop and ChromeOS by > moving it into a separate set of *MinidumpSymbolizer classes. > > Bug: chromium:994274 > Change-Id: I50cf0c0ff31f2f8934f51c2d342f73e9b405663a > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1884465 > Commit-Queue: Brian Sheedy <bsheedy@chromium.org> > Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Bug: chromium:994274 Change-Id: I5bd0510f5f258b9f4dd785c66e52f3102001e748 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1887677 Commit-Queue: Brian Sheedy <bsheedy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
These are no longer used. Bug: chromium:992427 Change-Id: I5a0dee8c76ec2c590d371013180d68fdacebcce5 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1879214 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Seems to be a name clash with a property also called storyTags. Bug: chromium:1019183 Change-Id: Ibd3bbd59f702c1a4de5bb538e7578f6e3b3b2b6c Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1887390 Reviewed-by: Dean Berris <dberris@chromium.org> Commit-Queue: Simon Hatch <simonhatch@chromium.org>
The debug logs include console errors, which are very userful for debugging. Bug: chromium:1015942 Change-Id: Ibdc06e01b5846cb7b5305cedf45939abeb1724c0 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1890871 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Deepanjan Roy <dproy@chromium.org>
Chromium binaries are causing dashboard tests to time out. After debugging with xvfb turned off, it appears that the entire browser process freezes for unclear reasons. The logs always mention getting a proxy config as the last message. This problem does not show up on chrome builds, so switching it back. Adding the --enable-features=ForceWebRequestProxyForTest flag as suggested by the network team to get around the earlier proxy issues we were seeing. Bug: chromium:1015942 Change-Id: I402fe99e1a80a039186dc21c658a8be9e4485349 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1891213 Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Commit-Queue: Deepanjan Roy <dproy@chromium.org>
This reverts commit d6f6ff7. Reason for revert: Metric computation failing on Windows. Original change's description: > [Telemetry] Stop doing metric computation. > > The absence of the 'histogram_dicts.json' artifact will let Results > Processor take over metric computation. > > Bug: chromium:981349 > Change-Id: I2ae04477c565b0005ff19b8d93b8f49c202f08ff > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1878290 > Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> > Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Bug: chromium:981349 Bug: chromium:1019888 Change-Id: Ibe4cc28c547a9776a0f5aacb51b17f17015cbf06 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1891692 Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
This CL converts the few remaining tests on "pages" to test on the more general "story" objects. At the same time, remaining tests no longer involve a real browser, thus testing the story_runner "unit" more closely. Final "system_stub"s in the module are also replaced with more generic mock objects. Bug: chromium:1008852 Change-Id: Ib363baa448727ccf33da06742435c04cc58e11fe Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1893221 Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Lazy initialization of bins causes a datarace in a multithreaded environment: one thread can start using the bins while another one is still initializing them. This CL initializes bins before returning the object to the caller eliminating this problem. Bug: chromium:1019888 Change-Id: Ic2461332626b750e0a266947e15e0b5ab3cd4209 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1893220 Commit-Queue: Mikhail Khokhlov <khokhlov@google.com> Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org>
Some cleanups possible after recent CLs that removed no longer used features. Bug: chromium:973837 Change-Id: I0217927ea973b0edae06dd19a03c4102ac78a3dd Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1887044 Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Bug: chromium:1005504 Change-Id: I53c085dec88bee8618879c304b5df9a688ee9c2e Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1895312 Reviewed-by: Ben Pastene <bpastene@chromium.org> Commit-Queue: John Budorick <jbudorick@chromium.org>
This will allow other tests to reuse these dummy stories, while also helping bring story_runner_unnittest under the 1K lines limit for the pylint "too-many-lines" warning. Bug: chromium:1008852 Change-Id: I827c87e79d550c5407857be24c96123f5f0b35f8 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1893223 Commit-Queue: Caleb Rouleau <crouleau@chromium.org> Auto-Submit: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
Auto generated using common/bin/update_chrome_reference_binaries.py. I was hitting some strange timeout issues on tests while working on another CL, which seem to go away after updating the reference builds. Change-Id: I0462c0ae563fb5798e65e011fba7993971409796 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1888636 Commit-Queue: Deepanjan Roy <dproy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org>
Consider copying src/foo/bar to target target/foo/bar. If target/foo/bar exists, then this results in target/foo/bar/bar. The right fix is to recursively compare directories. In the short term, disable the enable_push_sync path so that all directories are compared. BUG: 1020716 TEST=none Change-Id: I06c6893c5be36d9e4b16a5e91e53588f1a835c6e Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1895324 Reviewed-by: John Budorick <jbudorick@chromium.org> Commit-Queue: Scott Violet <sky@chromium.org>
This reverts commit 1ce01db. Reason for revert: Underlying root cause has been fixed. Original change's description: > Revert "[Telemetry] Stop doing metric computation." > > This reverts commit d6f6ff7. > > Reason for revert: Metric computation failing on Windows. > > Original change's description: > > [Telemetry] Stop doing metric computation. > > > > The absence of the 'histogram_dicts.json' artifact will let Results > > Processor take over metric computation. > > > > Bug: chromium:981349 > > Change-Id: I2ae04477c565b0005ff19b8d93b8f49c202f08ff > > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1878290 > > Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> > > Reviewed-by: Caleb Rouleau <crouleau@chromium.org> > > Bug: chromium:981349 > Bug: chromium:1019888 > Change-Id: Ibe4cc28c547a9776a0f5aacb51b17f17015cbf06 > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1891692 > Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> > Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> TBR=perezju@chromium.org,khokhlov@google.com # Not skipping CQ checks because original CL landed > 1 day ago. Bug: chromium:981349, chromium:1019888 Change-Id: I9bdbda6a97d60a6cef33b9b028362dd81551005c Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1897495 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
It's unfortunate when some component of dev server tests freezes and we have to wait two hour for the bot timeout to be reached to see the error. This CL adds: - timeout-sec and timeout-retries flags to run_dev_server_tests. Tests now have until timeout-sec to finish, and they can be retried a number of times. - Some timeouts for dashboard and tracing tests on build_steps. I have seen these tests time out recently. The timeout value is set to roughly max(duration on a passing run from 5 random samples of last 100 runs) * 2. Bug: chromium:1015942 Change-Id: Ie5aecae2f94f3f9c0d23b6a499fa3cd49a64df05 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1876876 Commit-Queue: Deepanjan Roy <dproy@chromium.org> Reviewed-by: Caleb Rouleau <crouleau@chromium.org> Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org>
This CL removes from Telemetry all code paths used to upload artifacts. This will cause results processor to take over and do the upload. Also clean up the code so the only remaining processing stage, trace serialization, is called from a single point in story_runner. Bug: chromium:981349 Change-Id: I12daa01c121410b7c62f8145598ffd075b38fe3d Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1897788 Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
The concept of "tir_label" (a timeline interaction record label) no longer exists in Telemetry. It was replaced with a simpler "grouping_label" which just allows grouping stories into meaningful groups. This CL updates tracing code to use the new name. Bug: chromium:974237 Change-Id: I483ceb6ced4d42c6fc78c291000b03d03fb42ee3 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1897787 Reviewed-by: Deepanjan Roy <dproy@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
This makes it harder to properly build URLs containing this testPaths. To get the correct path it forces one to double escape % sings. Bug: chromium:981349 Change-Id: I0fd1b83c88604c569ebb3676432b47e74e17bcb3 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1899432 Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Mikhail Khokhlov <khokhlov@google.com> Auto-Submit: Juan Antonio Navarro Pérez <perezju@chromium.org> Reviewed-by: Mikhail Khokhlov <khokhlov@google.com>
This reverts commit bce667b. Reason for revert: This will break here: https://cs.chromium.org/chromium/src/tools/perf/core/results_processor/formatters/json3_output.py?rcl=743250dc23f89ffabe9c9c26d185e0c2cd0e6771&l=47 Original change's description: > [Telemetry] Do not percent encode story names > > This makes it harder to properly build URLs containing this testPaths. > To get the correct path it forces one to double escape % sings. > > Bug: chromium:981349 > Change-Id: I0fd1b83c88604c569ebb3676432b47e74e17bcb3 > Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1899432 > Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org> > Commit-Queue: Mikhail Khokhlov <khokhlov@google.com> > Auto-Submit: Juan Antonio Navarro Pérez <perezju@chromium.org> > Reviewed-by: Mikhail Khokhlov <khokhlov@google.com> TBR=perezju@chromium.org,khokhlov@google.com Change-Id: I65b7c079cc3ab6a66e66149ee40192e9a3a4df32 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: chromium:981349 Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1899433 Reviewed-by: Juan Antonio Navarro Pérez <perezju@chromium.org> Commit-Queue: Juan Antonio Navarro Pérez <perezju@chromium.org>
Remove --no-legacy from `gcloud beta emulators datastore start` All legacy code is removed by datastore emulator according to https://critique.corp.google.com/review/273272781. Remove --no-legacy which breaks the tests. Change-Id: I560ae3264efd04cbe39e9bc4ced472c6454be6ea Reviewed-on: https://chromium-review.googlesource.com/c/catapult/+/1899230 Commit-Queue: Chenlin Fan <fancl@chromium.org> Reviewed-by: Dean Berris <dberris@chromium.org>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
13 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Merge catapult latest