[TRAFODION-3089] DatabaseMetaData.getIndexInfo not work well

[mashengchen]

it's hard to update statistic message each time row datas change. so current version does not support CARDINALITY.

[Traf-Jenkins]

Check Test Started: https://jenkins.esgyn.com/job/Check-PR-master/2742/

[Traf-Jenkins]

Test Passed. https://jenkins.esgyn.com/job/Check-PR-master/2742/

[selvaganesang]

It looks like the change made via #594 to make the squirel to work is being undone. Is that what you wanted?

[mashengchen]

no, actually i remove the sb_histograms , because the MD schema does not have that table, also if there does not do update stastisitc, the CARDINALITY will not correct.
i discuss with @venkat1m , he can explain more.

[DaveBirdsall]

@mashengchen, @selvaganesang, @venkat1m: Where does this change stand? Is it ready for merging? Does it require more work?

[Traf-Jenkins]

Check Test Started: https://jenkins.esgyn.com/job/Check-PR-master/2879/

[Traf-Jenkins]

Test Failed. https://jenkins.esgyn.com/job/Check-PR-master/2879/

[mashengchen]

jenkins,retest

[Traf-Jenkins]

Check Test Started: https://jenkins.esgyn.com/job/Check-PR-master/2944/

[Traf-Jenkins]

Test Passed. https://jenkins.esgyn.com/job/Check-PR-master/2944/

[moscowgentalman]

@mashengchen, @selvaganesang, @venkat1m: Where does this change stand? Is it ready for merging? Does it require more work?

[mashengchen]

@traflm this PR is for the bug which can get index in MD schema by the getIndexInfo api, but one can get result by using get indexes in trafci or sqlci

[DaveBirdsall]

@selvaganesang and @venkat1m: Do you have further comments? Or should I merge this?

[DaveBirdsall]

jenkins, retest

[Traf-Jenkins]

Check Test Started: https://jenkins.esgyn.com/job/Check-PR-master/3072/

[Traf-Jenkins]

Test Passed. https://jenkins.esgyn.com/job/Check-PR-master/3072/

[selvaganesang]

It looks like GetStatistics API was displaying some information about the table itself to take into account the primary key earlier. I am assuming that the primary key acts like an implied index that needs to be displayed as part of GetStatistics API.

Yes. It is incorrect to use sb_histograms table, but I fear that the row pertaining to the primary key is being completely removed from the output with this change. Hence can you please attach the output of the catalog API before and after your change to determine if this change can be merged. So my comment about #594 still stands good