Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Sqoop >> mail # dev >> Review Request: SQOOP-846 Provide Netezza connector


Copy link to this message
-
Re: Review Request: SQOOP-846 Provide Netezza connector


> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > Hi Venkat,
> > thank you very much for incorporating all my suggestions. I believe that we're almost there! I do have just couple of additional high level notes and few comments (mostly just nits):
> >
> > 1) Tests are passing for me - it was something specific to my environment. Thank you for checking though!
> >
> > 2) The docs are not compiling for me:
> >
> >      [exec] asciidoc: ERROR: connectors.txt: line 288: [blockdef-listing] missing closing delimiter
> >
> > 3) I would like to see more tests, especially about testing the various extra arguments. But I'm open to resolve that in follow up jira. No need to postpone this task even further.

Thanks Jarek for reviewing and feedback.

1)  Great - I could run it with NPS 6 and 7.  And glad that it is now passing for you.  
2)  Fixed the docs.
3)  I agree.   I have couple of manual tests but will integrate them by filing followon JIRAs

Will update new patch after testing and making sure the documentation is in order also

Thanks
Venkat
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/docs/user/connectors.txt, line 257
> > <https://reviews.apache.org/r/9543/diff/8/?file=263535#file263535line257>
> >
> >     Nit: Copy paste from PostgreSQL :-)

Thanks :)
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/docs/user/connectors.txt, lines 259-276
> > <https://reviews.apache.org/r/9543/diff/8/?file=263535#file263535line259>
> >
> >     It seems that the start of second column in the header is somewhere else than rest of the rows. This seems to be causing very weird typesetting of the table. Would you mind taking a look?

Yes.   Fixing it
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/docs/user/connectors.txt, line 262
> > <https://reviews.apache.org/r/9543/diff/8/?file=263535#file263535line262>
> >
> >     Nit: Would you mind prepending the argument with "\--"

I based on the Postgresql connector doc which was not prepending the --.  But adding it as it seems more natural for users to expect
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/docs/user/connectors.txt, lines 294-296
> > <https://reviews.apache.org/r/9543/diff/8/?file=263535#file263535line294>
> >
> >     Would you mind enclosing the execution example in "---"? It will be then typeset differently in the HTML docs.

Yes.  Good thing to do
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/java/org/apache/sqoop/manager/DirectNetezzaManager.java, lines 134-135
> > <https://reviews.apache.org/r/9543/diff/8/?file=263538#file263538line134>
> >
> >     Nit: Those two variables seems to be unused in the method.

Fixed.  It was on exportTable.
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/java/org/apache/sqoop/manager/DirectNetezzaManager.java, lines 141-142
> > <https://reviews.apache.org/r/9543/diff/8/?file=263538#file263538line141>
> >
> >     Nit: Return value of the methods is already char, is there a reason for the explicit retyping?

When I moved the check from execution to client side I did not change it where the method used was typing it to int
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/java/org/apache/sqoop/manager/NetezzaManager.java, line 96
> > <https://reviews.apache.org/r/9543/diff/8/?file=263540#file263540line96>
> >
> >     Shouldn't be safer operation to rollback instead of commit here? I would expect that the rest of code will commit any changes that it needed to do prior calling close method.

Yes.  That is true.  It seems that this is a carry over from the PostgresqlManager.   I will file a JIRA to fix the PostgresqlManager also
> On March 2, 2013, 7:26 p.m., Jarek Cecho wrote:
> > src/java/org/apache/sqoop/manager/NetezzaManager.java, line 223
> > <https://reviews.apache.org/r/9543/diff/8/?file=263540#file263540line223>
> >
> >     I found this quite confusing. The if(cmdLine.hasOption()) do not have else and thus that path of the execution will leave this property unset. It seems that the default value is false (per NetezzaDataDrivenDBInputFormat). I would suggest to either set this property in all branches or depend on the default value.

Sorry about the confusion.  I cleaned it up and removed the else clause with the default value set unconditionally and optionally reset

Good point.   Changing it to be consistent

Yes.  Makes sense.
- Venkat
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/9543/#review17303
On Feb. 28, 2013, 8:06 a.m., Venkat Ranganathan wrote: