impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dan Hecht (Code Review)" <>
Subject [Impala-ASF-CR] IMPALA-4134,IMPALA-3704: Kudu INSERT improvements
Date Fri, 21 Oct 2016 21:41:52 GMT
Dan Hecht has posted comments on this change.

Change subject: IMPALA-4134,IMPALA-3704: Kudu INSERT improvements

Patch Set 6:

File be/src/exec/

PS6, Line 145: .
why?  (that's the more interesting part for the code reader since the code tells us the "what"
part as well). 

also, what's the benefit of the user specifying a larger buffer if we're not going to fill

Line 282:                      << "previous write operation might be inconsistent.\n";
can't the write operation be inconsistent even if there was no overflow? i.e. can't it be
inconsistent if there was a single error?

Line 298:         error_msg_buffer << "Kudu error(s) reported, first error: " <<
why not just create the Status here since we no longer will append to it later? and you can
just use status.ok() and get rid of first_error/failed.

PS6, Line 330: kudu_error_counter_->value()
this doesn't really work in case of GetPendingErrors() returning overflow, right? are we okay
with ignoring that case?
File be/src/exec/kudu-table-sink.h:

PS6, Line 48: due to a 
can't it return an error for other reasons?

PS6, Line 49: be reported as warnings.
isn't the first Kudu error also reported as the error status?

PS6, Line 113: if the Kudu client may be
             :   /// getting behind.
what does this mean? does the code use it, or is this saying a user could use it?  and how
can that be determined from this timer?
File common/thrift/

PS6, Line 300: Error Kudu table 
should this say "Error in Kudu table ..." or something?

To view, visit
To unsubscribe, visit

Gerrit-MessageType: comment
Gerrit-Change-Id: I5542b9a061b01c543a139e8722560b1365f06595
Gerrit-PatchSet: 6
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Matthew Jacobs <>
Gerrit-Reviewer: Alex Behm <>
Gerrit-Reviewer: Dan Hecht <>
Gerrit-Reviewer: Matthew Jacobs <>
Gerrit-Reviewer: Todd Lipcon <>
Gerrit-HasComments: Yes

View raw message