Marcel Kornacker has posted comments on this change.

Change subject: IMPALA-3481: Use Kudu ScanToken API for scan ranges
......................................................................


Patch Set 4:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/4120/4/fe/src/main/java/com/cloudera/impala/planner/KuduScanNode.java
File fe/src/main/java/com/cloudera/impala/planner/KuduScanNode.java:

Line 131:    * TODO: Remove when the columns are fetched from Kudu directly 
during analysis.
> when you "open" a table, it fetches the metadata (columns, etc) from the ma
one of our design principles is to avoid round-trips to central servers during 
plan generation. (i believe your test cluster numbers, but i also believe it's 
not that hard to create an adversarial workload.)

let's remove this todo.


Line 139:         tableSchema.getColumn(colName);
> We're more or less following the pattern of JDBC where trying to get an inv
or why not simply have getColumn() return null?

we have no special allegiance to jdbc.


-- 
To view, visit http://gerrit.cloudera.org:8080/4120
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I160e5849d372755748ff5ba3c90a4651c804b220
Gerrit-PatchSet: 4
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Matthew Jacobs <[email protected]>
Gerrit-Reviewer: Dan Burkert <[email protected]>
Gerrit-Reviewer: Dimitris Tsirogiannis <[email protected]>
Gerrit-Reviewer: Marcel Kornacker <[email protected]>
Gerrit-Reviewer: Matthew Jacobs <[email protected]>
Gerrit-Reviewer: Todd Lipcon <[email protected]>
Gerrit-HasComments: Yes

Reply via email to