Closed longquanzheng closed 3 years ago
Changes Missing Coverage | Covered Lines | Changed/Added Lines | % | ||
---|---|---|---|---|---|
internal/internal_event_handlers.go | 3 | 5 | 60.0% | ||
<!-- | Total: | 4 | 6 | 66.67% | --> |
Files with Coverage Reduction | New Missed Lines | % | ||
---|---|---|---|---|
internal/internal_task_pollers.go | 8 | 80.67% | ||
<!-- | Total: | 8 | --> |
Totals | |
---|---|
Change from base Build 45d338ee-a2b0-4411-b30a-df23bc7d684a: | -0.03% |
Covered Lines: | 12167 |
Relevant Lines: | 17005 |
What changed? Use originalRunID to fix childWF decision panic after reset due to runID change
Why? For https://github.com/uber-go/cadence-client/issues/1045 Use originalRunID is the right fix for this, and is backward-compatible. https://github.com/uber/cadence-idl/blob/48a14966426238c15570a13deab5a7905c0a89c1/thrift/shared.thrift#L556
See more discussion in https://uber-cadence.slack.com/archives/CLS1SM941/p1607515148008700 There are a couple of users running into this issues.
How did you test it? Asking a user to test it.
Potential risks No, it should be backward compatible. Without reset, orignalRunID should be the same as runID.