Adar Dembo has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/11753 )

Change subject: authz: verify tokens on scans
......................................................................


Patch Set 8:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/11753/7/src/kudu/common/schema.h
File src/kudu/common/schema.h:

http://gerrit.cloudera.org:8080/#/c/11753/7/src/kudu/common/schema.h@552
PS7, Line 552:   inline const std::vector<ColumnId>& column_ids() const {
> Done
Isn't inline implied by virtue of this being a function defined in the header? 
Like, columns() doesn't need inline, so why should this?


http://gerrit.cloudera.org:8080/#/c/11753/4/src/kudu/tserver/tablet_service.cc
File src/kudu/tserver/tablet_service.cc:

http://gerrit.cloudera.org:8080/#/c/11753/4/src/kudu/tserver/tablet_service.cc@427
PS4, Line 427:       return false;
> I think there are a couple of approaches worth considering:
If a user sends raw RPCs in order to get a diff scan (i.e. two timestamps) in 
UNORDERED mode, they may see multiple ghost entries for the same PK in the 
output. They'll need to deal with that somehow, either by deduplicating them 
client-side, or trying to order them chronologically, though the server doesn't 
return enough information to accomplish this.

Agreed that we might want to enforce that if a diff scan was requested, ORDERED 
mode was also requested. No such enforcement exists today: 
https://gerrit.cloudera.org/c/12592

As a strawman, I'd propose that virtual columns require privileges on PK 
columns. Why? Because many (if not all) of the virtual columns I can think of 
can be viewed as bits of metadata on specific rows, so you need to at least 
establish a row's identity in order to use them effectively (i.e. to know which 
row we're talking about). And you need PK access for that.



--
To view, visit http://gerrit.cloudera.org:8080/11753
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I7a5d81cf215a5d936f8853feba05778038764905
Gerrit-Change-Number: 11753
Gerrit-PatchSet: 8
Gerrit-Owner: Andrew Wong <[email protected]>
Gerrit-Reviewer: Adar Dembo <[email protected]>
Gerrit-Reviewer: Andrew Wong <[email protected]>
Gerrit-Reviewer: Dan Burkert <[email protected]>
Gerrit-Reviewer: Hao Hao <[email protected]>
Gerrit-Reviewer: Kudu Jenkins (120)
Gerrit-Reviewer: Tidy Bot (241)
Gerrit-Comment-Date: Mon, 18 Mar 2019 17:48:11 +0000
Gerrit-HasComments: Yes

Reply via email to