Skip to content

Tags: ibmsoe/kudu

Tags

1.1.0

Toggle 1.1.0's commit message
Release Apache Kudu 1.1.0

1.1.0-RC1

Toggle 1.1.0-RC1's commit message
Release note updates for 1.1

Change-Id: Ifa6bf79830d0b7fd2595d1ba64ea3673e90e5008
Reviewed-on: http://gerrit.cloudera.org:8080/5079
Reviewed-by: Jean-Daniel Cryans <[email protected]>
Reviewed-by: Adar Dembo <[email protected]>
Tested-by: Kudu Jenkins
Tested-by: Todd Lipcon <[email protected]>
(cherry picked from commit b46c4ec)
Reviewed-on: http://gerrit.cloudera.org:8080/5081
Reviewed-by: Todd Lipcon <[email protected]>

1.0.1

Toggle 1.0.1's commit message

Verified

This tag was signed with the committer’s verified signature.
danburkert Dan Burkert
Release Apache Kudu 0.10.0

1.0.1-RC1

Toggle 1.0.1-RC1's commit message
Bump version to 1.0.1

Change-Id: I3a8804f076593b0a49f1ad42771e734a45ab222e
Reviewed-on: http://gerrit.cloudera.org:8080/4667
Tested-by: Dan Burkert <[email protected]>
Reviewed-by: Adar Dembo <[email protected]>

1.0.0

Toggle 1.0.0's commit message
Release Apache Kudu 1.0.0

1.0.0-RC1

Toggle 1.0.0-RC1's commit message
[client] avoid circular deps in time-based flusher

The boost::bind() makes cast of parameters during the call,
not during creation of the functor object:
  http://stackoverflow.com/questions/11255144/why-does-boostbind-store-arguments-of-the-type-passed-in-rather-than-of-the-ty

So, it's necessary to pass weak pointers to the background auto-flush
task to avoid circular dependencies between client::KuduSession::Data
and rpc::Messenger.  Besides, it does not make much sense to store
shared reference to messenger in KuduSession::Data since it's always
passed as a weak reference and then promoting to a shared one during
the call in all usage scenarios.

Thanks to Adar and Todd spotting the usual suspect there.

This is a follow-up for 93be131.

Change-Id: I59825981a600f5882ee476479c2ddf16b495c1f9
Reviewed-on: http://gerrit.cloudera.org:8080/4395
Reviewed-by: Adar Dembo <[email protected]>
Tested-by: Kudu Jenkins
Tested-by: Todd Lipcon <[email protected]>
(cherry picked from commit 1a06225)
Reviewed-on: http://gerrit.cloudera.org:8080/4403
Reviewed-by: Todd Lipcon <[email protected]>

0.10.0

Toggle 0.10.0's commit message
Release Apache Kudu 0.10.0

0.10.0-RC1

Toggle 0.10.0-RC1's commit message
Enable replay cache by default

We've done looping of the stress tests, as well as some significant
cluster testing (YCSB on 72 node cluster for several days). In the
cluster, we didn't see any significant increased memory usage due to
this feature, nor any stability issues.

Additionally, the risk of the feature is relatively low: the worst
it could cause is memory leaks or crashes, but no chance for permanent
data corruption or loss. For the time being, I am leaving the ability
to disable the feature via a gflag, in case we do see any instability.
We can remove the flag entirely in the 1.0 timeframe.

Change-Id: I35b5e74ac30aa3309a0a7e035c8dff7f61c3f275
Reviewed-on: http://gerrit.cloudera.org:8080/4003
Reviewed-by: Todd Lipcon <[email protected]>
Tested-by: Kudu Jenkins

0.9.1

Toggle 0.9.1's commit message
Release Apache Kudu (incubating) 0.9.1

0.9.1-RC1

Toggle 0.9.1-RC1's commit message
KUDU-1385 Part 1: Modify documentation landing page

Rename introduction.html to index.html and get rid
of old index.html  completely.

Change-Id: Icb4b5b5ba4851d9a7974b413f8637089a157973a
Reviewed-on: http://gerrit.cloudera.org:8080/2675
Tested-by: Kudu Jenkins
Reviewed-by: Todd Lipcon <[email protected]>
(cherry picked from commit 738e993)