| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
|
How is this PR related to #711 ? It seems they both server the same purpose, no? |
Sorry, something went wrong.
The fix for the bug is implemented in the PR #711. The fix for the bug is implemented in the same way as in the findJdbcExecutionOperatorTaskInStage() method to fix the bug instead of WayangCollections.getSingle(). This PR is similar to the previous one in the sense that it is another fix for the robustness bug of JdbcExecutor. The only difference is the logic used to fix the bug in this PR is different from the previous one. The previous one could cause a bug due to the assumption made for the consumers. The reason for choosing to submit two PRs is that it is easier to review them separately. If you think the two PRs should be merged into one, I am happy to merge them. |
Sorry, something went wrong.
| Back | FazBrowse Home | New Git URL |
This PR makes the JdbcExecutor more robust by removing the faulty use of
WayangCollections.getSingle() in the function findJdbcExecutionOperatorTaskInStage().
The faulty function could potentially throw an exception if there were zero or
multiple consumers on the channel. This could cause the following runtime errors:
"[] is not a singleton"
This change will guarantee that the function now checks the number of consumers
before trying to access them, returning null if it doesn’t match the execution
pattern.
This is a status quo change, removing unneeded runtime errors in the JDBC
execution path.