[
https://issues.jboss.org/browse/TEIID-5700?page=com.atlassian.jira.plugin...
]
Steven Hawkins updated TEIID-5700:
----------------------------------
Description:
The search being used to find a max ndv for the independent side is just setting the
result to be the original ndv estimate for the independent side. This results in
unnecessary back-offs of performing the dependent join with a log message similar to:
TEIID30011 Not performing dependent join using source $dsc/id0, since the number of
distinct rows for expression g0.fkIdProductCode exceeds 1.333. You should ensure that
your source statistics, including column distinct value counts, accurately reflect the
source or use a MAKE_DEP hint to force the join.
was:The search being used to find a max ndv for the independent side is just setting the
result to be the original ndv estimate for the independent side. This results in
unnecessary back-offs of performing the dependent join.
Dependent join back-off max ndv set to independent ndv estimate
---------------------------------------------------------------
Key: TEIID-5700
URL:
https://issues.jboss.org/browse/TEIID-5700
Project: Teiid
Issue Type: Bug
Components: Query Engine
Reporter: Steven Hawkins
Assignee: Steven Hawkins
Priority: Major
Fix For: 12.2, 11.2.3, 12.1.1
The search being used to find a max ndv for the independent side is just setting the
result to be the original ndv estimate for the independent side. This results in
unnecessary back-offs of performing the dependent join with a log message similar to:
TEIID30011 Not performing dependent join using source $dsc/id0, since the number of
distinct rows for expression g0.fkIdProductCode exceeds 1.333. You should ensure that
your source statistics, including column distinct value counts, accurately reflect the
source or use a MAKE_DEP hint to force the join.
--
This message was sent by Atlassian Jira
(v7.12.1#712002)