On Sat, Feb 21, 2015 at 2:45 AM, Enis Söztutar enis@gmail.com wrote:
git tag has this directory that is not in source tarball:
Only in .: hbase-native-client
This is interesting. We've never added this to the release tarball it
seems. Given that there is no usage (not sure why we
* the source tarball does not match the git tag (ignoring dot files).
source tarball has these files that are not in the git tag:
Only in /Users/busbey/Downloads/hbase-1.0.0/dev-support:
test-patch.properties.orig
Only in /Users/busbey/Downloads/hbase-1.0.0/dev-support: test-patch.sh.orig
Thanks for everybody who tested the release and voted.
The vote passes with 7 +1's (6 binding). I'll push out the artifacts now
and the send a release notes Sunday.
Enis
On Fri, Feb 20, 2015 at 11:05 PM, Jeffrey Zhong jzh...@hortonworks.com
wrote:
+1. I have compiled run Phoenix master
Thanks for tests and votes so far.
Here is my official +1 (see above for my tests).
I was inspecting a similar issue today in the cluster with
IntegrationTestLoadAndVerify with destructive CM, and the log splitting did
not finish in 3 attempts. I think the issue is a not that critical though,
I
See HBASE-13070, I found the problem finally. It is a testcase issue so I
do not think it should block a release.
2015-02-20 10:12 GMT+08:00 Andrew Purtell apurt...@apache.org:
Maybe its just me
On Thu, Feb 19, 2015 at 11:10 AM, Enis Söztutar enis@gmail.com
wrote:
I did not run the
+1
Checked signatures
Ran test suite - passed
Ran basic shell commands - passed
Compiled Hive against maven artifacts.
Cheers
On Sat, Feb 14, 2015 at 9:55 PM, Enis Söztutar e...@apache.org wrote:
It gives me great pleasure to announce that the sixth release candidate for
the release
1.0.0
+1
- Checked sums and signature
- Source and binary tarball layouts look ok
- Spot checked docs
- Built from source
- All unit tests pass (7u75)
- Most integration tests pass in the verification phase of hbase-it (I
failed to pass -DskipITs to the build),
except IntegrationTestIngestWithACL,
+1
Checked signature
Checked source and binary tars
All unit tests passed on JDK 7
ITBLL worked and passed.
On Fri, Feb 20, 2015 at 4:22 PM, Andrew Purtell apurt...@apache.org wrote:
+1
- Checked sums and signature
- Source and binary tarball layouts look ok
- Spot checked docs
- Built
+1 (non-binding)
* verified src signature (using http://www.apache.org/dist/hbase/KEYS)
* verified licenses (with apache-rat:check and by inspecting LICENSE.txt
and NOTICE.txt)
* built Apache Hive against RC artifacts and ran through the HBase Handler
qtests (works post HIVE-9706, currently
+1. I have compiled run Phoenix master branch code(with some code
changes) on top of Hbase 1.0.0. All Phoenix tests are passed and one issue
is found(https://issues.apache.org/jira/browse/HBASE-13077)
Thanks,
-Jeffrey
On 2/20/15, 12:38 AM, Enis Söztutar e...@apache.org wrote:
Thanks for tests
I did not run the unit tests locally. I was relying on the following build
which is 100% pass:
https://builds.apache.org/view/All/job/HBase-1.0/746/
It has the exact RC5 bits (6c98bff7b719efdb16f71606f3b7d8229445eb81). Let
me kick a local UT run just in case.
Enis
On Thu, Feb 19, 2015 at 9:05
That list is missing unit testing. Did you happen to do that when building
but not mention it Enis? If so did they all pass for you?
On Wed, Feb 18, 2015 at 7:48 PM, Enis Söztutar enis@gmail.com wrote:
Here is my RC5 testing so far:
- checked checksums, sigs
- checked the bin and src
Maybe its just me
On Thu, Feb 19, 2015 at 11:10 AM, Enis Söztutar enis@gmail.com wrote:
I did not run the unit tests locally. I was relying on the following build
which is 100% pass:
https://builds.apache.org/view/All/job/HBase-1.0/746/
It has the exact RC5 bits
+1
+ Sig and MD5 verified.
+ Built from src tgz successfully.
+ Started in standalone mode. Did some clickey-clickey. All worked.
+ Loaded a bit of data and verified it loaded. Nothing untoward in logs.
+ Layout looks good.
+ Built bin tarball. Looks good.
+ Built site. Javadoc, doc., and site
Thanks J-M for doing these tests.
Some comments inlined.
On Mon, Feb 16, 2015 at 7:37 AM, Jean-Marc Spaggiari
jean-m...@spaggiari.org wrote:
Download and un-packed passed.
Checked Changes.txt = Passed.
Checked documentation = Link Why does HBase care about /etc/hosts?
Comments inlined too ;)
Download and un-packed passed.
Checked Changes.txt = Passed.
Checked documentation = Link Why does HBase care about /etc/hosts?
http://devving.com/?p=414 in section Quick Start - Standalone HBase
doesn't work
Seems like a minor issue.
Even smaller than
I'm not able to get a clean unit test run when building from source using
7u67. TestCacheOnWrite and TestSplitLogManager fail for me, maybe more but
the build doesn't get past hbase-server. Maybe these are known issues? If
not I'll dig in when I get some time.
On Sat, Feb 14, 2015 at 9:55 PM,
TestCacheOnWrite itself has some problems. It
uses TestHFileWriterV2.randomOrderedKey to generate a random byte array,
then use first 32 bytes as row and other parts as family and qualifier, but
TestHFileWriterV2.randomOrderedKey may return a byte array only contains 32
bytes, so there will be
Interesting. I just got a clean build on jdk1.7.0_67 (Failed in ITs). Will
retry a few times.
St.Ack
On Wed, Feb 18, 2015 at 3:11 PM, Andrew Purtell apurt...@apache.org wrote:
I'm not able to get a clean unit test run when building from source using
7u67. TestCacheOnWrite and
I remember both of these as flaky tests from the past, but I am not sure
about their current status.
https://builds.apache.org/view/All/job/HBase-1.0/ does not have any of
these two failures for the recent builds.
Enis
On Wed, Feb 18, 2015 at 3:11 PM, Andrew Purtell apurt...@apache.org wrote:
Here is my RC5 testing so far:
- checked checksums, sigs
- checked the bin and src artifacts
- checked layouts
- checked java files in src tarball, and jar files in bin tarball
- checked the book and the site (they are new style)
- checked javadocs for both devapi and userapi
- checked
Quick question, how do you validate the signature? Seems to be a compressed
format, not sure if there is a specific command to validate it.
You can use:
gpg --recv-key RM_KEY
gpg artifact.asc
to verify the sig. https://www.apache.org/dev/release-signing.html contains
all the details.
Hum. When I do that I get a warning saying that the key might not belong to
its owner... But expect this waning, I'm able to validate the
hbase-1.0.0-bin.tar.gz file.
Thanks,
JM
2015-02-17 14:25 GMT-05:00 Enis Söztutar enis@gmail.com:
Quick question, how do you validate the signature?
No, 1,0.0 to 1.0.0 ;) Installed 1.0.0, changed a config in a hbase-site.xml
and did a rolling restart to have it taken into consideration.
I still have migration from 0.94 to 1.0.0 to test, and rolling restart from
0.9x to 1.0.0... And of course, the performances...
JM
2015-02-16 10:40
Hi Enis,
Quick question, how do you validate the signature? Seems to be a compressed
format, not sure if there is a specific command to validate it.
Thanks,
JM
2015-02-15 0:55 GMT-05:00 Enis Söztutar e...@apache.org:
It gives me great pleasure to announce that the sixth release candidate for
Download and un-packed passed.
Checked Changes.txt = Passed.
Checked documentation = Link Why does HBase care about /etc/hosts?
http://devving.com/?p=414 in section Quick Start - Standalone HBase
doesn't work
Run test suite = Failed 3 time in a row with JDK 1.7
Failed tests:
bq. Did a rolling restart from 1.0.0 to 1.0.0
Did you mean from 0.98 to 1.0.0 ?
Cheers
On Mon, Feb 16, 2015 at 7:37 AM, Jean-Marc Spaggiari
jean-m...@spaggiari.org wrote:
Download and un-packed passed.
Checked Changes.txt = Passed.
Checked documentation = Link Why does HBase care about
It gives me great pleasure to announce that the sixth release candidate for
the release
1.0.0 (HBase-1.0.0RC5), is available for download at
https://dist.apache.org/repos/dist/dev/hbase/hbase-1.0.0RC5/
Maven artifacts are also available in the temporary repository
28 matches
Mail list logo
