fix: failure to commit leads to inconsistencies between workflow and database

[rbeuque74]

• What kind of change does this PR introduce? (Bug fix, feature, docs update, ...) Bug Fix

• What is the current behavior? (You can also link to an open issue here)

When a µTask worker fails to commit its current state to the database, it was sometimes not stopping the execution of its resolution. This leads to inconsistencies, as the step was currently RUNNING, but database state indicates TODO. In that way, if the worker crashes, then the new worker will restart with wrong information, inducing workflow issues.

• Does this PR introduce a breaking change? (What changes might users need to make in their application due to this PR?) No

• Other information:

[ovh-cds]

CDS Report test#1183.0 ✘

• tests and lint
  • golangci-lint ✘
  • unit-test and integration-tests ✔

[ovh-cds]

CDS Report package#1184.0 ✘

• Stage 1
  • Docker build ✘