Thanks for this really heavy lift as RM, Stack. Simple thanks doesn't do it
justice really, the scope of this release has been super demanding, but you
have them nonetheless.
On Sun, Apr 29, 2018 at 10:13 PM, Stack wrote:
> With 4 non-binding +1s, 4 binding +1s, two -0s, and a
With 4 non-binding +1s, 4 binding +1s, two -0s, and a +0, the vote passes.
We branched hbase-2 four years ago. Our first 2.0.0 alpha was almost a year
ago. Though there are perf issues still (see reports on HBASE-20483,
currently, actively been worked-on) and some interesting late arrivals
+1
S
On Sun, Apr 22, 2018 at 10:10 PM, Stack wrote:
> The third release candidate for Apache HBase 2.0.0 is available for
> downloading and testing.
>
> Artifacts are available here:
>
> https://dist.apache.org/repos/dist/dev/hbase/hbase-2.0.0RC2/
>
> Maven artifacts are
On Fri, Apr 27, 2018 at 3:06 PM, 张铎(Duo Zhang)
wrote:
> For master and branch-2 it will break serial replication feature, on
> branch-2.0, I do not think it will cause over-play but it may cause
> something wrong with region locating(not sure, no evidence), as we use the
>
Stack, HBASE-20504 has been filed for this issue. Thanks.
Sakthi
On Fri, Apr 27, 2018 at 2:37 PM, Stack wrote:
> On Fri, Apr 27, 2018 at 2:21 PM, Azhaku Sakthi Vel Muthu Krishnan <
> jatsak...@cloudera.com> wrote:
>
> > Stack, it works well for CentOS(7.2), Mac OS(10.13.3)
For master and branch-2 it will break serial replication feature, on
branch-2.0, I do not think it will cause over-play but it may cause
something wrong with region locating(not sure, no evidence), as we use the
open sequence number to determine whether we should replace the cache.
2018-04-28
On Fri, Apr 27, 2018 at 2:21 PM, Azhaku Sakthi Vel Muthu Krishnan <
jatsak...@cloudera.com> wrote:
> Stack, it works well for CentOS(7.2), Mac OS(10.13.3) & Ubuntu(16.04). It
> showed up the same issue on another machine with CentOS(7.4 i.e. same as
> before). So, my best guess, it's specific
Andrew, the java version was same while testing for Mac(10.13.3) &
CentOS(7.4). It worked on the mac though.
Sakthi
On Fri, Apr 27, 2018 at 1:26 PM, Andrew Purtell wrote:
> TestHeapSize failures are more likely JVM related.
>
>
> On Fri, Apr 27, 2018 at 12:36 PM, Stack
Stack, it works well for CentOS(7.2), Mac OS(10.13.3) & Ubuntu(16.04). It
showed up the same issue on another machine with CentOS(7.4 i.e. same as
before). So, my best guess, it's specific centos version(7.4) dependent?
Sakthi
*Sakthi* | Software Engineer
t. (480) 335-3681 <00>
TestHeapSize failures are more likely JVM related.
On Fri, Apr 27, 2018 at 12:36 PM, Stack wrote:
> On Fri, Apr 27, 2018 at 11:54 AM, Azhaku Sakthi Vel Muthu Krishnan <
> jatsak...@cloudera.com> wrote:
>
> > +0
> >
> > - signatures and checksums: OK
> > - rat check: OK
> > -
On Fri, Apr 27, 2018 at 11:54 AM, Azhaku Sakthi Vel Muthu Krishnan <
jatsak...@cloudera.com> wrote:
> +0
>
> - signatures and checksums: OK
> - rat check: OK
> - built from source (1.8.0_171): OK
> - ran Unit Tests
>
> NOTE:
>
> The following 3 tests fail when Unit Tests are attempted on a CentOS
+0
- signatures and checksums: OK
- rat check: OK
- built from source (1.8.0_171): OK
- ran Unit Tests
NOTE:
The following 3 tests fail when Unit Tests are attempted on a CentOS
platform.
TestHeapSize.testMutations:520 expected:<120> but was:<112>
TestHeapSize.testNativeSizes:200 expected:<48>
On Fri, Apr 27, 2018 at 3:54 AM, 张铎(Duo Zhang)
wrote:
> -0
>
> See https://issues.apache.org/jira/browse/HBASE-20476
>
> There is a race between reportRegionTransition and regionServerReport which
> could lead to openSeqNum go backwards. But the sequence id at RS side is
>
-0
See https://issues.apache.org/jira/browse/HBASE-20476
There is a race between reportRegionTransition and regionServerReport which
could lead to openSeqNum go backwards. But the sequence id at RS side is
correct so I'm not sure whether it is worth to sink a RC.
2018-04-26 15:59 GMT+08:00
+1 (non-binding)
Checksum, signature: ok
Rat check: ok
Build from source (8_163): ok
Start from bin in pseudo distributed mode w/ HDFS: ok
LTT 1M rows: ok
Simple shell commands: ok
Test RSGroup: ok
Web UI: ok
REST server: ok
NOTE: Using the new JDK (8_171) TestSaslFanOutOneBlockAsyncDFSOutput
+1
Checked signatures : good
Ran test suite with Java 8 : passed
> On Wed, Apr 25, 2018 at 1:12 PM, Andrew Purtell
> wrote:
>
> > +1
> >
> > Checked sums and signatures: ok
> > Built from source: ok (8u152)
> > RAT check passes: ok (8u152)
> > LTT 1M rows 20% updates: ok
Thanks Lars.
S
On Wed, Apr 25, 2018 at 1:28 AM, Lars Francke
wrote:
> Thanks Stack.
>
> In that case I'm -0 on this RC.
> I'd really like to either revert that patch or at least document the
> deprecation so we can remove it in 3.0. The way it is now we can't remove
>
Thanks Stack.
In that case I'm -0 on this RC.
I'd really like to either revert that patch or at least document the
deprecation so we can remove it in 3.0. The way it is now we can't remove
getTypeByte() in 3.0 because we'll need it for getType() which we can't
change because it's not documented
+1
Checked sums and signatures: ok
Built from source: ok (8u152)
RAT check passes: ok (8u152)
LTT 1M rows 20% updates: ok (8u152)
ITBLL 100M rows: ok (8u152)
On Sun, Apr 22, 2018 at 10:10 PM, Stack wrote:
> The third release candidate for Apache HBase 2.0.0 is available for
+1 (binding)
On 4/23/18 1:10 AM, Stack wrote:
The third release candidate for Apache HBase 2.0.0 is available for
downloading and testing.
Artifacts are available here:
https://dist.apache.org/repos/dist/dev/hbase/hbase-2.0.0RC2/
Maven artifacts are available in the staging repository at:
+1 non-binding
download src & bin tar ball - OK
signatures & sums - OK
build from source (openjdk version "1.8.0_151") - OK
rat check-
OK
unit
On Tue, Apr 24, 2018 at 1:25 AM, Lars Francke
wrote:
> Stack,
>
> I'm sure you'll want to slap me for bringing this up so late (I only found
> it recently) but could you take a look at the recent discussions in <
> https://issues.apache.org/jira/browse/HBASE-19746>?
> Is
+1
signature: pass.
build: pass.
PE (randomWrite/randomRead): pass.
zkcli, wal, hfile, regionsplitter tools: pass.
started Thrift server: pass.
started REST server: pass.
--
Cloudera, Inc.
On Tue, Apr 24, 2018 at 3:13 AM, Balazs Meszaros <
balazs.mesza...@cloudera.com> wrote:
> +1
>
> -
Hi all,
I'm sorry, I didn't mean to send this mail as an answer to this thread but
separate. It does not belong in this discussion.
Cheers,
Lars
On Tue, Apr 24, 2018 at 4:25 PM, Lars Francke
wrote:
> Stack,
>
> I'm sure you'll want to slap me for bringing this up so
Stack,
I'm sure you'll want to slap me for bringing this up so late (I only found
it recently) but could you take a look at the recent discussions in <
https://issues.apache.org/jira/browse/HBASE-19746>?
Is there any need for me to work on this or would it be too late to merge
this change anyway?
+1
- signatures and checksums: OK
- rat check: OK
- built from source (8u112): OK
- basic shell functionalities: OK
- web UI: OK
- LTT with 1M rows: OK
On Mon, Apr 23, 2018 at 7:10 AM, Stack wrote:
> The third release candidate for Apache HBase 2.0.0 is available for
>
The third release candidate for Apache HBase 2.0.0 is available for
downloading and testing.
Artifacts are available here:
https://dist.apache.org/repos/dist/dev/hbase/hbase-2.0.0RC2/
Maven artifacts are available in the staging repository at:
27 matches
Mail list logo