Github user ijokarumawak commented on a diff in the pull request:
https://github.com/apache/nifi/pull/3179#discussion_r235274647
--- Diff:
nifi-nar-bundles/nifi-hive-bundle/nifi-hive3-processors/src/main/java/org/apache/nifi/processors/hive/PutHive3QL.java
---
@@ -148,7 +148,23 @@ public void constructProcess() {
if (e instanceof SQLNonTransientException) {
return ErrorTypes.InvalidInput;
} else if (e instanceof SQLException) {
- return ErrorTypes.TemporalFailure;
+ // Use the SQLException's vendor code for guidance -- see
Hive's ErrorMsg class for details on error codes
+ int errorCode = ((SQLException) e).getErrorCode();
+ if (errorCode >= 10000 && errorCode < 20000) {
+ return ErrorTypes.InvalidInput;
+ } else if (errorCode >= 20000 && errorCode < 30000) {
+ return ErrorTypes.TemporalFailure;
--- End diff --
Based on Hive source code and the detail of these errors, I think we should
map 2xxxx error code to InvalidInput as retrying will not succeed.
> 20000 to 29999: Runtime errors where Hive believes that retries are
unlikely to succeed.
https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/ErrorMsg.java#L48
---
