Race condition when multiple people update an issue.

Issue #71 new
Jaiden Ashmore created an issue

This is a theoretical problem and I haven't seen it but it is possible to happen and therefore should be considered. However, maybe it is so rare that it never happens.

Two people update an issue a the same time. Current updaters = [anotherUser]

  1. "firstUser" updates the issues.
  2. (Thread 1) Microservice receives first webHook.
  3. (Thread 1) Gets the changelog and processes the updaters as [anotherUser, firstUser]
  4. "secondUser" updates the issues.
  5. (Thread 2) A different microservice receives second webHook.
  6. (Thread 2) Gets the changelog and processes the updaters as [anotherUser, firstUser, secondUser]
  7. (Thread 2) Posts to JIRA with this updater list [anotherUser, firstUser, secondUser]
  8. (Thread 2) Done.
  9. (Thread 1) Post to JIRA with this updater list [anotherUser, firstUser].
  10. (Thread 1) Done.
  11. We now have a state where the updaters is not correct.

This is more prevalent when dealing the lastUpdater search criteria.

Comments (0)

  1. Log in to comment