Skip to content

Conversation

@self-transition
Copy link
Member

@self-transition self-transition commented Mar 12, 2025

  1. Cause of requestsCache didn't lock listWatch onResponse callback, In some cases, multiple connections may have been established and the older connections are not managed by the global store.
  2. In previous fix _isDestroyed has too great a reach, The flag should only affect the get method

Before
We can see that two connections were established when the same resource was initialized, and that the connection established first was taken out of the management

iShot_2025-03-12_13 38 31

After
iShot_2025-03-12_13 41 11

BTW, there is an another case left: if the request onResponse callback is timeout and didn't trigger error, the cache request lock will not be removed

@self-transition self-transition added the bug Something isn't working label Mar 12, 2025
@self-transition self-transition merged commit 8dfff21 into main Mar 17, 2025
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants