Todd Lipcon has posted comments on this change.

Change subject: KUDU-2119. Fix failure in encoding-test
......................................................................


Patch Set 3:

(3 comments)

http://gerrit.cloudera.org:8080/#/c/7967/2/src/kudu/cfile/encoding-test.cc
File src/kudu/cfile/encoding-test.cc:

PS2, Line 271: int len = local_rng.Uniform(8)
> What if len == 0 as a result of calling local_rng.Uniform(8)?  Is that the 
yea I think that's fine and desirable since it tests zero-length strings. The 
"1 byte per entry" is still valid since we expect there to be at least a 
one-byte length for each entry, even if it's zero, in the current string 
encodings. (perhaps in the future we would encode lengths more efficiently but 
not today)


http://gerrit.cloudera.org:8080/#/c/7967/3/src/kudu/cfile/encoding-test.cc
File src/kudu/cfile/encoding-test.cc:

Line 279:     const uint kCount = r.Uniform(1000) + 1;
> Why need to '+1' here? Is it to ensure to add at least one element to the b
will do


PS3, Line 285: kCount
> Is it possible for s.size() == kCount? Can it be exactly 1 byte per entry?
given there is some footer/header stuff as well, we're guaranteed more than one 
byte per


-- 
To view, visit http://gerrit.cloudera.org:8080/7967
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Ic46d2a993235e560475d931c7b023eb5b4faf7c2
Gerrit-PatchSet: 3
Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-Owner: Todd Lipcon <[email protected]>
Gerrit-Reviewer: Adar Dembo <[email protected]>
Gerrit-Reviewer: Alexey Serbin <[email protected]>
Gerrit-Reviewer: Andrew Wong <[email protected]>
Gerrit-Reviewer: Hao Hao <[email protected]>
Gerrit-Reviewer: Kudu Jenkins
Gerrit-Reviewer: Tidy Bot
Gerrit-Reviewer: Todd Lipcon <[email protected]>
Gerrit-Reviewer: Will Berkeley <[email protected]>
Gerrit-HasComments: Yes

Reply via email to