Bug #72246 | executeBatch insert count broken by rewriteBatchedStatements=true | ||
---|---|---|---|
Submitted: | 4 Apr 2014 22:13 | Modified: | 9 Apr 2014 7:45 |
Reporter: | Justin Cranford | Email Updates: | |
Status: | Duplicate | Impact on me: | |
Category: | Connector / J | Severity: | S3 (Non-critical) |
Version: | 5.1.58-1~dotdeb.0-log (Debian) | OS: | Linux (Debian x64 v5.0.5 (Lenny)) |
Assigned to: | Assigned Account | CPU Architecture: | Any |
[4 Apr 2014 22:13]
Justin Cranford
[4 Apr 2014 22:15]
Justin Cranford
Correction: Instead of getting 1 back for each element of the returned int[] array, all of the values are *NOW* N.
[7 Apr 2014 6:59]
Alexander Soklakov
Hi Justin, If you give us more info about errors, maybe provide stacktrace, we could be more precise in our analysis. According to given information this problem might be the same Cody Lerum reported for Hibernate where compliance was broken by Bug#72246 patch, return values greater than 2 (insert ... on duplicate) are not accepted by some frameworks. So the proposal here, and patch in progress in related Bug#61213, is to use JDBC compliant value Statement.SUCCESS_NO_INFO instead of total count per each rewritten statement result.
[7 Apr 2014 7:02]
Alexander Soklakov
UPD: "...where compliance was broken by Bug#68562 patch..."
[7 Apr 2014 12:49]
Justin Cranford
I have no stack trace to post beyond the error logged by my example code. The error I get is specific to my application. Connector/J does not throw an exception, it is proprietary application logic that logged the error.
[9 Apr 2014 7:45]
Alexander Soklakov
Justin, I close this report as duplicate of Bug#68562. Please, feel free to reopen it if you find that your situation is different.
[9 Apr 2014 7:53]
Alexander Soklakov
Please note that fix will be introduced in bounds of Bug#61213