Hello Dan Burkert, Mike Percy, Todd Lipcon,
I'd like you to do a code review. Please visit
to review the following change.
Change subject: KUDU-687: expose additional tablet metadata in C++ client
KUDU-687: expose additional tablet metadata in C++ client
This patch adds new data-only KuduReplica and KuduTablet classes.
Along with KuduTabletServer, the C++ client now has more parity with the
Java client w.r.t. exposing tablet metadata.
For now, this is only exposed via KuduScanToken, because it's already doing
the work to figure out which replicas exist where. That's an odd fit for
ksck (which doesn't scan, at least not using the client), but it should
stave off any controversy stemming from adding dubious public APIs. In the
future, it wouldn't be unreasonable for these classes to be exposed via
KuduTable in some way.
There are four public signature changes:
- Addition of KuduScanToken::tablet(): this is backwards compatible.
- Addition of KuduTabletServer::port(): this is backwards compatible.
- Change to the KuduScanToken constructor: this should be backwards
compatible, because it's private and so shouldn't used by applications.
- Removal of KuduScanToken::TabletServers(): this is an incompatible change.
I think it's OK because Impala is the only significant C++ client user and
it's not even using scan tokens yet.
15 files changed, 500 insertions(+), 95 deletions(-)
git pull ssh://gerrit.cloudera.org:29418/kudu refs/changes/46/4146/1
To view, visit http://gerrit.cloudera.org:8080/4146
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings
Gerrit-Owner: Adar Dembo <a...@cloudera.com>
Gerrit-Reviewer: Dan Burkert <d...@cloudera.com>
Gerrit-Reviewer: Mike Percy <mpe...@apache.org>
Gerrit-Reviewer: Todd Lipcon <t...@apache.org>