084719cde1
I can't say with certainty yet what caused the two failures I know about, but I have one theory: the node continuing to check in during the test skewed the online score towards 1, and using the test default for GracefulExitDurationInDays meant there were fewer update periods than expected. At any rate, it is more correct to pause the graceful exit processing chore and the contact chore during the test, even if it doesn't end up solving the problem. Refs: https://github.com/storj/storj/issues/6401 Change-Id: I06d43d531e0b3344af13878c8d55213349fdcfa3 |
||
|---|---|---|
| .. | ||
| chore_test.go | ||
| common.go | ||
| db_test.go | ||
| db.go | ||
| endpoint_test.go | ||
| endpoint.go | ||
| getnodes_test.go | ||
| gracefulexit_test.go | ||
| observer_test.go | ||
| observer.go | ||
| pending_test.go | ||
| pending.go | ||
| validation.go | ||