Massive performance degradation in getCacheKey()
See original GitHub issueš Bug Report
The getCacheKey()
function seems to degradate as the amount of files increases. I observed this in a project with a few thousands of files when upgrading from Jest 26 to 27.
To Reproduce
A few hundreds of files which have to be transformed.
Expected behavior
Performance remains constant.
Link to repo (highly encouraged)
Unfortunately I canāt share the repo. But I added some logging to that function.
jest@27.0.5 ts-jest@27.0.3
# node_modules/ts-jest/dist/ts-jest-transformer.js
var i = 0;
// ...
TsJestTransformer.prototype.getCacheKey = function (fileContent, filePath, transformOptions) {
var start = new Date().getTime();
var _a;
var configs = this._configsFor(transformOptions);
this._logger.debug({ fileName: filePath, transformOptions: transformOptions }, 'computing cache key for', filePath);
var _b = transformOptions.instrument, instrument = _b === void 0 ? false : _b;
var constructingCacheKeyElements = [
this._transformCfgStr,
exports.CACHE_KEY_EL_SEPARATOR,
configs.rootDir,
exports.CACHE_KEY_EL_SEPARATOR,
"instrument:" + (instrument ? 'on' : 'off'),
exports.CACHE_KEY_EL_SEPARATOR,
fileContent,
exports.CACHE_KEY_EL_SEPARATOR,
filePath,
];
if (!configs.isolatedModules && this._tsResolvedModulesCachePath) {
var resolvedModuleNames = void 0;
if (((_a = this._depGraphs.get(filePath)) === null || _a === void 0 ? void 0 : _a.fileContent) === fileContent) {
this._logger.debug({ fileName: filePath, transformOptions: transformOptions }, 'getting resolved modules from disk caching or memory caching for', filePath);
resolvedModuleNames = this._depGraphs
.get(filePath)
.resolvedModuleNames.filter(function (moduleName) { return fs_1.existsSync(moduleName); });
}
else {
this._logger.debug({ fileName: filePath, transformOptions: transformOptions }, 'getting resolved modules from TypeScript API for', filePath);
resolvedModuleNames = this._compiler.getResolvedModules(fileContent, filePath, transformOptions.cacheFS);
this._depGraphs.set(filePath, {
fileContent: fileContent,
resolvedModuleNames: resolvedModuleNames,
});
fs_1.writeFileSync(this._tsResolvedModulesCachePath, json_1.stringify(__spreadArray([], __read(this._depGraphs))));
}
resolvedModuleNames.forEach(function (moduleName) {
constructingCacheKeyElements.push(exports.CACHE_KEY_EL_SEPARATOR, moduleName, exports.CACHE_KEY_EL_SEPARATOR, fs_1.statSync(moduleName).mtimeMs.toString());
});
}
var end = new Date().getTime();
console.error('_getCacheKey #' + i + ' ' + (end - start));
i++
return sha1_1.sha1.apply(void 0, __spreadArray([], __read(constructingCacheKeyElements)));
};
# Start
_getCacheKey #0 5503
_getCacheKey #1 0
_getCacheKey #2 12
_getCacheKey #3 9
_getCacheKey #4 42
# It's getting worse
_getCacheKey #500 18
_getCacheKey #501 21
_getCacheKey #502 18
_getCacheKey #503 19
_getCacheKey #504 18
# And worse (until all memory is used and system freezes)
_getCacheKey #1000 39
_getCacheKey #1001 39
_getCacheKey #1002 39
_getCacheKey #1003 42
_getCacheKey #1004 45
With Jest 26 everything is fine:
jest@26.6.3 ts-jest@26.5.6
# node_modules/ts-jest/dist/ts-jest-transformer.js
var i = 0;
// ...
TsJestTransformer.prototype.getCacheKey = function (fileContent, filePath, _jestConfigStr, transformOptions) {
var start = new Date().getTime();
var configs = this.configsFor(transformOptions.config);
this.logger.debug({ fileName: filePath, transformOptions: transformOptions }, 'computing cache key for', filePath);
var _a = transformOptions.instrument, instrument = _a === void 0 ? false : _a, _b = transformOptions.rootDir, rootDir = _b === void 0 ? configs.rootDir : _b;
var end = new Date().getTime();
console.error('_getCacheKey #' + i + ' ' + (end - start));
i++
return sha1_1.sha1(this._transformCfgStr, '\x00', rootDir, '\x00', "instrument:" + (instrument ? 'on' : 'off'), '\x00', fileContent, '\x00', filePath);
};
# Start
_getCacheKey #0 104
_getCacheKey #1 0
_getCacheKey #2 0
_getCacheKey #3 0
_getCacheKey #4 0
# Still fast
_getCacheKey #500 0
_getCacheKey #501 0
_getCacheKey #502 0
_getCacheKey #503 0
_getCacheKey #504 0
# No degradation after 1000 files
_getCacheKey #1000 0
_getCacheKey #1001 0
_getCacheKey #1002 0
_getCacheKey #1003 0
_getCacheKey #1004 0
envinfo
System:
OS: Ubuntu 18.04.1
Node version: v14.15.5
Npm packages:
jest: 27.0.5
ts-jest: 27.0.3
typescript: 4.2.4
babel(optional): 7.14.6
Issue Analytics
- State:
- Created 2 years ago
- Reactions:6
- Comments:8
Top Results From Across the Web
Work for Performance - Open performance-issues in open ...
a list of open performance-issues in most used open-source software. ... kulshekhar/ts-jest - Massive performance degradation in getCacheKey() ...
Read more >How We Improved the Performance of B2B eCommerce ...
Performance degradation is possible compared to in-memory cache due to ... a GetCacheKey () method that can be used to generate cache keys....
Read more >@jest/create-cache-key-function | Yarn - Package Manager
objectContaining() description (#6754); [babel-jest] Make getCacheKey() take into account createTransformer options (#6699); [jest-jasmine2] Use prettierĀ ...
Read more >Massive Performance Degradation On Statistics For Large ...
Oracle Database - Enterprise Edition - Version 19.3.0.0.0 and later: Massive Performance Degradation On Statistics For Large PartitionedĀ ...
Read more >Clearing the Cache for Selected Resources - MODX
In the default MODX setup, it gets the documents cache key with $doc->getCacheKey() , then tells cacheManager to delete the item with thatĀ ......
Read more >Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start FreeTop Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Top GitHub Comments
I logged #2707 and closely-related #1747. In the first, I added a comment that I had eventually implemented a workaround for that bug. Maybe this design change should be rolled back for the sake of performance in the short term. Itās a choice of which bug is worse.
Experiencing the same issue. We have 300+ Angular tests and I canāt even run them locally after migration.
ts-jest
consumes 60 GB of memory (16 RAM + 45 swap) and my macOS just crashes. Rollback tojest@26
fixes the issue.