Stress test/new tests #24936
Merged
Stress test/new tests #24936
Conversation
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
|
This pull request is automatically built and testable in CodeSandbox. To see build info of the built libraries, click here or the icon next to each commit SHA. Latest deployment of this branch, based on commit b21cfbe:
|
Collaborator
📊 Bundle size reportUnchanged fixtures
|
Asset size changesSize Auditor did not detect a change in bundle size for any component! Baseline commit: 42acccd4445b42c806f957de00e5f2567589279f (build) |
spmonahan
force-pushed
the
stress-test/new-tests
branch
from
74c1ee3 to
abe4d04
Compare
Adds tests cases for - mounting - injecting styles - adding lots of nodes - removing all nodes - re-rendering the entire page for Fluent V8, V9 and Web Components. Introduces the concept of "renderers" that easily allow different components to be slotting into tests with just a query parameter. Moves RandomTree code to scripts to handle a ts-node compilation issue. This isn't ideal but it doesn't seem worth spending lots of time figuring the proper issue out right now. Similarly, some of the randomization code has been moved around in scripts because ts-node was having difficulty importing it.
Adds a "target size" option to RandomTree that, when set, builds a tree to a specific size. Tweaks the default fixture parameters based on manual testing. Fixtures essentially double in size for each t-shirt size. Reduces the default number of Tachometer samples from 25 to 10. In some testing this still gives statistically significant results for all default sizes while making tests take significantly less time to run.
This feature was removed in favor of query params.
Fix inaccuracies, document new features.
spmonahan
force-pushed
the
stress-test/new-tests
branch
from
adab364 to
0a2bb73
Compare
github-actions
bot
modified the milestones:
September Project Cycle Q1 2022,
October Project Cycle Q4 2022
spmonahan
force-pushed
the
stress-test/new-tests
branch
from
0a2bb73 to
a08591f
Compare
Collaborator
Perf Analysis (
|
| Scenario | Render type | Master Ticks | PR Ticks | Iterations | Status |
|---|---|---|---|---|---|
| Avatar | mount | 1316 | 1342 | 5000 | |
| Button | mount | 930 | 953 | 5000 | |
| FluentProvider | mount | 1584 | 1589 | 5000 | |
| FluentProviderWithTheme | mount | 633 | 635 | 10 | |
| FluentProviderWithTheme | virtual-rerender | 595 | 598 | 10 | |
| FluentProviderWithTheme | virtual-rerender-with-unmount | 635 | 609 | 10 | |
| MakeStyles | mount | 1937 | 1892 | 50000 | |
| SpinButton | mount | 2607 | 2533 | 5000 |
Collaborator
Perf Analysis (
|
| Scenario | Current PR Ticks | Baseline Ticks | Ratio |
|---|---|---|---|
| ButtonMinimalPerf.default | 175 | 157 | 1.11:1 |
| AlertMinimalPerf.default | 265 | 242 | 1.1:1 |
| VideoMinimalPerf.default | 741 | 680 | 1.09:1 |
| ButtonSlotsPerf.default | 532 | 492 | 1.08:1 |
| ChatDuplicateMessagesPerf.default | 261 | 242 | 1.08:1 |
| IconMinimalPerf.default | 651 | 609 | 1.07:1 |
| RadioGroupMinimalPerf.default | 439 | 416 | 1.06:1 |
| AttachmentSlotsPerf.default | 1100 | 1048 | 1.05:1 |
| LabelMinimalPerf.default | 374 | 357 | 1.05:1 |
| FlexMinimalPerf.default | 275 | 265 | 1.04:1 |
| GridMinimalPerf.default | 329 | 317 | 1.04:1 |
| ReactionMinimalPerf.default | 366 | 352 | 1.04:1 |
| TextMinimalPerf.default | 340 | 326 | 1.04:1 |
| DropdownManyItemsPerf.default | 644 | 628 | 1.03:1 |
| HeaderMinimalPerf.default | 349 | 339 | 1.03:1 |
| HeaderSlotsPerf.default | 753 | 732 | 1.03:1 |
| ListNestedPerf.default | 551 | 535 | 1.03:1 |
| SegmentMinimalPerf.default | 343 | 334 | 1.03:1 |
| SkeletonMinimalPerf.default | 328 | 319 | 1.03:1 |
| AccordionMinimalPerf.default | 142 | 139 | 1.02:1 |
| AttachmentMinimalPerf.default | 143 | 140 | 1.02:1 |
| BoxMinimalPerf.default | 336 | 331 | 1.02:1 |
| LayoutMinimalPerf.default | 353 | 345 | 1.02:1 |
| MenuMinimalPerf.default | 832 | 815 | 1.02:1 |
| SliderMinimalPerf.default | 1600 | 1572 | 1.02:1 |
| StatusMinimalPerf.default | 667 | 652 | 1.02:1 |
| TreeMinimalPerf.default | 804 | 789 | 1.02:1 |
| CarouselMinimalPerf.default | 443 | 439 | 1.01:1 |
| DatepickerMinimalPerf.default | 5717 | 5658 | 1.01:1 |
| DialogMinimalPerf.default | 759 | 754 | 1.01:1 |
| DividerMinimalPerf.default | 341 | 339 | 1.01:1 |
| FormMinimalPerf.default | 363 | 360 | 1.01:1 |
| ImageMinimalPerf.default | 375 | 373 | 1.01:1 |
| InputMinimalPerf.default | 1114 | 1104 | 1.01:1 |
| PopupMinimalPerf.default | 620 | 614 | 1.01:1 |
| ProviderMergeThemesPerf.default | 1270 | 1254 | 1.01:1 |
| TableMinimalPerf.default | 394 | 391 | 1.01:1 |
| AnimationMinimalPerf.default | 520 | 522 | 1:1 |
| AvatarMinimalPerf.default | 186 | 186 | 1:1 |
| CardMinimalPerf.default | 520 | 520 | 1:1 |
| ChatMinimalPerf.default | 703 | 700 | 1:1 |
| DropdownMinimalPerf.default | 2663 | 2668 | 1:1 |
| EmbedMinimalPerf.default | 3633 | 3639 | 1:1 |
| ListMinimalPerf.default | 496 | 496 | 1:1 |
| RosterPerf.default | 2118 | 2117 | 1:1 |
| ProviderMinimalPerf.default | 401 | 401 | 1:1 |
| RefMinimalPerf.default | 216 | 216 | 1:1 |
| SplitButtonMinimalPerf.default | 4341 | 4333 | 1:1 |
| TableManyItemsPerf.default | 1851 | 1851 | 1:1 |
| TooltipMinimalPerf.default | 2249 | 2250 | 1:1 |
| ItemLayoutMinimalPerf.default | 1160 | 1173 | 0.99:1 |
| ListCommonPerf.default | 608 | 615 | 0.99:1 |
| MenuButtonMinimalPerf.default | 1668 | 1687 | 0.99:1 |
| CustomToolbarPrototype.default | 2630 | 2663 | 0.99:1 |
| ButtonOverridesMissPerf.default | 1283 | 1312 | 0.98:1 |
| CheckboxMinimalPerf.default | 2063 | 2100 | 0.98:1 |
| ListWith60ListItems.default | 585 | 597 | 0.98:1 |
| LoaderMinimalPerf.default | 661 | 673 | 0.98:1 |
| ToolbarMinimalPerf.default | 882 | 898 | 0.98:1 |
| TreeWith60ListItems.default | 154 | 158 | 0.97:1 |
| PortalMinimalPerf.default | 160 | 166 | 0.96:1 |
| ChatWithPopoverPerf.default | 351 | 369 | 0.95:1 |
| TextAreaMinimalPerf.default | 431 | 458 | 0.94:1 |
Collaborator
Perf Analysis (
|
| Scenario | Render type | Master Ticks | PR Ticks | Iterations | Status |
|---|---|---|---|---|---|
| BaseButton | mount | 1179 | 1170 | 5000 | |
| Breadcrumb | mount | 2928 | 2957 | 1000 | |
| Checkbox | mount | 2652 | 2630 | 5000 | |
| CheckboxBase | mount | 2357 | 2323 | 5000 | |
| ChoiceGroup | mount | 4368 | 4392 | 5000 | |
| ComboBox | mount | 1239 | 1273 | 1000 | |
| CommandBar | mount | 9575 | 9578 | 1000 | |
| ContextualMenu | mount | 11208 | 11355 | 1000 | |
| DefaultButton | mount | 1357 | 1368 | 5000 | |
| DetailsRow | mount | 3606 | 3605 | 5000 | |
| DetailsRowFast | mount | 3641 | 3594 | 5000 | |
| DetailsRowNoStyles | mount | 3418 | 3513 | 5000 | |
| Dialog | mount | 3103 | 3102 | 1000 | |
| DocumentCardTitle | mount | 575 | 571 | 1000 | |
| Dropdown | mount | 3200 | 3203 | 5000 | |
| FocusTrapZone | mount | 2007 | 1988 | 5000 | |
| FocusZone | mount | 1959 | 2009 | 5000 | |
| GroupedList | mount | 52198 | 58622 | 2 | |
| GroupedList | virtual-rerender | 24827 | 24455 | 2 | |
| GroupedList | virtual-rerender-with-unmount | 95956 | 92807 | 2 | |
| GroupedListV2 | mount | 560 | 557 | 2 | |
| GroupedListV2 | virtual-rerender | 522 | 535 | 2 | |
| GroupedListV2 | virtual-rerender-with-unmount | 553 | 570 | 2 | |
| IconButton | mount | 1865 | 1905 | 5000 | |
| Label | mount | 729 | 735 | 5000 | |
| Layer | mount | 4266 | 4218 | 5000 | |
| Link | mount | 822 | 820 | 5000 | |
| MenuButton | mount | 1677 | 1673 | 5000 | |
| MessageBar | mount | 2368 | 2364 | 5000 | |
| Nav | mount | 3263 | 3279 | 1000 | |
| OverflowSet | mount | 1353 | 1361 | 5000 | |
| Panel | mount | 2545 | 2523 | 1000 | |
| Persona | mount | 1238 | 1282 | 1000 | |
| Pivot | mount | 1666 | 1647 | 1000 | |
| PrimaryButton | mount | 1507 | 1524 | 5000 | |
| Rating | mount | 7015 | 7071 | 5000 | |
| SearchBox | mount | 1534 | 1486 | 5000 | |
| Shimmer | mount | 2808 | 2840 | 5000 | |
| Slider | mount | 2086 | 2067 | 5000 | |
| SpinButton | mount | 4702 | 4653 | 5000 | |
| Spinner | mount | 795 | 794 | 5000 | |
| SplitButton | mount | 3102 | 3109 | 5000 | |
| Stack | mount | 858 | 858 | 5000 | |
| StackWithIntrinsicChildren | mount | 2348 | 2348 | 5000 | |
| StackWithTextChildren | mount | 4801 | 4785 | 5000 | |
| SwatchColorPicker | mount | 10526 | 10337 | 5000 | |
| TagPicker | mount | 2626 | 2698 | 5000 | |
| TeachingBubble | mount | 86243 | 87690 | 5000 | |
| Text | mount | 778 | 770 | 5000 | |
| TextField | mount | 1593 | 1593 | 5000 | |
| ThemeProvider | mount | 1522 | 1522 | 5000 | |
| ThemeProvider | virtual-rerender | 1090 | 1075 | 5000 | |
| ThemeProvider | virtual-rerender-with-unmount | 2137 | 2174 | 5000 | |
| Toggle | mount | 1097 | 1109 | 5000 | |
| buttonNative | mount | 533 | 545 | 5000 |
Improve test consistency across browsers. Firefox and Chrome/Edge can measure different things in certain scenarios so this commit updates the measurements for mount to account for this. The React Profiler has been enabled so we can get timing information from React. In particular this is useful for knowing when React started doing work.
spmonahan
force-pushed
the
stress-test/new-tests
branch
from
a08591f to
e0cbefb
Compare
Hotell
reviewed
Oct 11, 2022
|
|
||
| terser@^4.1.2, terser@^4.3.9, terser@^4.6.12, terser@^4.6.2, terser@^4.6.3: | ||
| version "4.8.1" | ||
| version "4.8.0" |
Contributor
There was a problem hiding this comment.
wondering what triggered this downgrade ?
Contributor
Author
There was a problem hiding this comment.
Not sure since the only dep change I made was to remove chromedriver and it does not depend on Terser. I've also pulled in upstream changes. Maybe something from there?
Hotell
approved these changes
Oct 11, 2022
Contributor
Hotell
left a comment
Hotell
left a comment
There was a problem hiding this comment.
approving for v-build affected changes
micahgodbolt
approved these changes
Oct 11, 2022
JustSlone
approved these changes
Oct 11, 2022
Collaborator
JustSlone
left a comment
JustSlone
left a comment
There was a problem hiding this comment.
I see this is a large PR, I assume this is a one-time thing given the specific phase a development the test was in, and, in the future, we'll have more specific PRs for particular changes.
Tachometer sometimes fails during a test run. It's not entirely clear why this happens but it seems like there are two scenarios: 1. chromedriver needs to be installed or updated. Tachometer handles this dependency for us but when the dep is installed the test run fails. Subsequent test runs are fine so retry logic addresses this. 2. Timing is off. This issue is less clear but it seems like when a browser is slow to start Tachometer might time out. This doesn't seem like an actual issue with the tool and more just occasional flakiness in the system. While not addressing the root cause of either of these issues retry gives stress-test the ability to handle little speed bumps better.
NotWoods
pushed a commit
to NotWoods/fluentui
that referenced
this pull request
Nov 18, 2022
* stress test: adds test cases for v8, v9 and web components Adds tests cases for - mounting - injecting styles - adding lots of nodes - removing all nodes - re-rendering the entire page for Fluent V8, V9 and Web Components. Introduces the concept of "renderers" that easily allow different components to be slotting into tests with just a query parameter. Moves RandomTree code to scripts to handle a ts-node compilation issue. This isn't ideal but it doesn't seem worth spending lots of time figuring the proper issue out right now. Similarly, some of the randomization code has been moved around in scripts because ts-node was having difficulty importing it. * stress test: configure fixtures, tree generation and sample sizes Adds a "target size" option to RandomTree that, when set, builds a tree to a specific size. Tweaks the default fixture parameters based on manual testing. Fixtures essentially double in size for each t-shirt size. Reduces the default number of Tachometer samples from 25 to 10. In some testing this still gives statistically significant results for all default sizes while making tests take significantly less time to run. * stress test: remove 'renderers' cli option This feature was removed in favor of query params. * stress test: update readme Fix inaccuracies, document new features. * stress-test: improve tests Improve test consistency across browsers. Firefox and Chrome/Edge can measure different things in certain scenarios so this commit updates the measurements for mount to account for this. The React Profiler has been enabled so we can get timing information from React. In particular this is useful for knowing when React started doing work. * Change default sample-size from 10 to 25 * add stress-test to codeowners * stress-test: add retry logic for tachometer Tachometer sometimes fails during a test run. It's not entirely clear why this happens but it seems like there are two scenarios: 1. chromedriver needs to be installed or updated. Tachometer handles this dependency for us but when the dep is installed the test run fails. Subsequent test runs are fine so retry logic addresses this. 2. Timing is off. This issue is less clear but it seems like when a browser is slow to start Tachometer might time out. This doesn't seem like an actual issue with the tool and more just occasional flakiness in the system. While not addressing the root cause of either of these issues retry gives stress-test the ability to handle little speed bumps better.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 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.
Current Behavior
We don't have any actual test cases.
New Behavior
Reduces test sample size from 25 to 10. This still gives statistically significant results but takes much less time to run.Related Issue(s)
Partially fixes #24424