Add saveCacheV2 tests
#1879
Add saveCacheV2 tests
#1879
Conversation
|
|
||
| test('save with missing input should fail', async () => { | ||
| const paths: string[] = [] | ||
| const primaryKey = 'Linux-node-bb828da54c148048dd17899ba9fda624811cfb43' |
There was a problem hiding this comment.
nit: to be consistent with the restoreCacheV2 tests, how about we name this key?
| }) | ||
|
|
||
| test('save with large cache outputs should fail using', async () => { | ||
| const filePath = 'node_modules' |
There was a problem hiding this comment.
nit: also here, maybe it's better to use paths instead?
| expect(createCacheEntryMock).toHaveBeenCalledTimes(1) | ||
| expect(createCacheEntryMock).toHaveBeenCalledWith({ | ||
| key: primaryKey, | ||
| version: cacheVersion | ||
| }) |
There was a problem hiding this comment.
Thinking out loud: are both of these necessary? Wouldn't one be sufficient?
| const createCacheEntryMock = jest | ||
| .spyOn(CacheServiceClientJSON.prototype, 'CreateCacheEntry') | ||
| .mockReturnValue(Promise.resolve({ok: false, signedUploadUrl: ''})) |
There was a problem hiding this comment.
As I'm reading the test, I'm thinking whether it might be better to shuffle the code, try to call CreateCacheEntry earlier and only create the archive if this call succeeds. Right now it seems we're doing a lot of work on the runner (creating the archive first) before checking whether we can create the cache entry
| const filePath = 'node_modules' | ||
| const primaryKey = 'Linux-node-bb828da54c148048dd17899ba9fda624811cfb43' | ||
| const logWarningMock = jest.spyOn(core, 'warning') | ||
| const signedUploadURL = 'https://signed-upload-url.com' |
There was a problem hiding this comment.
nit: do we want to use the same url?
Fixes #1879
restoreCacheV2tests #1876, this PR adds unit tests for thesaveCacheV2uploadCacheFileand includes a debug message with the status returned on upload