hudi-agent commented on code in PR #18805:
URL: https://github.com/apache/hudi/pull/18805#discussion_r3282061604
##########
hudi-hadoop-common/src/main/java/org/apache/hudi/common/config/DFSPropertiesConfiguration.java:
##########
@@ -156,8 +156,8 @@ public void addPropsFromFile(StoragePath filePath) {
);
try {
- if (filePath.equals(DEFAULT_PATH) && !storage.exists(filePath)) {
- log.debug("Properties file {} not found. Ignoring to load props file",
filePath);
+ if (!storage.exists(filePath)) {
Review Comment:
🤖 This now silently tolerates a missing file for *every* caller, including
the top-level constructor path used by `UtilHelpers.readConfig`,
`StreamerUtil.readConfig`, and `HDFSParquetImporterUtils.readConfig` (Delta
Streamer `--props`, Flink streamer, Spark CLI). Previously a typo in `--props
/missing.props` produced a clear `HoodieIOException`; after this change the job
silently runs with an empty `TypedProperties` and just a WARN log — which can
lead to confusing downstream failures or jobs running with unintended defaults.
Have you considered scoping the relaxed behavior to the `include=` recursion
case (e.g. an internal overload with an `isInclude` flag) so explicit
user-provided top-level paths still fail fast? The added tests only cover
`include=` scenarios; a missing top-level path isn't exercised. @yihua /
@nsivabalan WDYT?
<sub><i>- AI-generated; verify before applying. React 👍/👎 to flag
quality.</i></sub>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
