chore: split cypress files for less memory

apache / superset

Apache Superset is a Data Visualization and Data Exploration Platform

https://superset.apache.org/

Apache License 2.0

61.99k stars 13.59k forks source link

chore: split cypress files for less memory #30354

Closed eschutho closed 2 days ago

eschutho commented 1 week ago

SUMMARY

Testing to see if this will fix the cypress memory issues. Also added the chrome flag disable-dev-shm-usage

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

TESTING INSTRUCTIONS

ADDITIONAL INFORMATION

[ ] Has associated issue:
[ ] Required feature flags:
[ ] Changes UI
[ ] Includes DB Migration (follow approval process in SIP-59)
- [ ] Migration is atomic, supports rollback & is backwards-compatible
- [ ] Confirm DB migration upgrade and downgrade tested
- [ ] Runtime estimates and downtime expectations provided
[ ] Introduces new feature or API
[ ] Removes existing feature or API

mistercrunch commented 4 days ago

Matrix (1, chrome) failed so I restarted it manually. For the record dashboard/nativeFilters.test.ts is the one that failed, even after shrinking it down as part of this PR. I'm wondering if lowering the limit / cardinality of filters could help here too if the issue comes from too many filters with too high of a cardinality (assuming that this specific test is the hot spot).

Wondering if a wider matrix would help. Also it seems we have 5 * chrome now, but could try running also against Firefox to see if the memory consuption is more or less of a problem on that browser.

rusackas commented 3 days ago

Matrix (1, chrome) failed so I restarted it manually. For the record dashboard/nativeFilters.test.ts is the one that failed, even after shrinking it down as part of this PR. I'm wondering if lowering the limit / cardinality of filters could help here too if the issue comes from too many filters with too high of a cardinality (assuming that this specific test is the hot spot).

Wondering if a wider matrix would help. Also it seems we have 5 * chrome now, but could try running also against Firefox to see if the memory consuption is more or less of a problem on that browser.

Yes, that is the hot spot, and it's always the one that fails. We've also seen GitHub issues about performance regressions when there are lots of filters, so this failure is likely reflecting a real problem. We can throw memory, split tests, reduce test scope, etc. to smooth out devEx, but I think we're just dancing around a real problem.

eschutho commented 3 days ago

@rusackas this is passing CI.. I ran it again and had no errors.. can we try to merge it and test it out on master?

mistercrunch commented 2 days ago

Looked at master hoping to see some green but no :/

mistercrunch commented 2 days ago

Giving a shot at troubleshooting memory here -> https://github.com/apache/superset/pull/30397 , seems promising