[BUG] 'Cannot read from socket/invalid date' error

[filcuk]

Describe the bug ~~Sudden error stops scheduler from triggering further tests, as well as manual tests.~~ I haven't found any other faults on the server. Works perfectly fine after a restart.

Update: It does not stop the scheduler, the single test failed as well as the manual one I ran a day later, but the rest of the scheduled tests all ran successfully.

To Reproduce Steps to reproduce the behavior:

1. Not known, will update if I find out

Expected behavior Service should run a test as scheduled.

Screenshots The first error below is the scheduled test. Second error is me attempting to run a manual test.

2022-01-22T21:14:59.641572182Z [services.d] starting services
2022-01-22T21:14:59.668598469Z [services.d] done.
2022-01-25T15:00:09.087654505Z {"type":"log","timestamp":"2022-01-25T15:00:09Z","message":"Error: [11] Cannot read from socket: Resource temporarily unavailable","level":"error"}
2022-01-26T00:00:09.455555406Z {"type":"log","timestamp":"2022-01-26T00:00:09Z","message":"Error: [11] Cannot read: Resource temporarily unavailable","level":"error"}
2022-01-26T17:23:28.866648130Z [cont-finish.d] executing container finish scripts...
2022-01-26T17:23:28.867931346Z [cont-finish.d] done.

Context

• Using volume storage

[filcuk]

I've just found the test in time of the error shows Invalid date in results:

[z0r3f]

Does it relate to #459 ? Same pattern for me and I solved it.

[PhilippeSch]

This seems to rather be a duplicate of #725 where the failed tests occur randomly...

[filcuk]

Checked #725, hard to say if it's the same issue. Wouldn't think it's GDPR related, as the one test failed, but other scheduled tests run successfully, after which the next manual test failed again. My TZ is Europe/London

Only two errors in ~200 tests so far, will update if I have more data.