Looking at
https://builds.apache.org/view/All/job/Phoenix-4.4-HBase-1.1/2/testReport/org.apache.phoenix.end2end/GroupByCaseIT/org_apache_phoenix_end2end_GroupByCaseIT/
:
Caused by: org.apache.hadoop.hbase.DoNotRetryIOException:
org.apache.hadoop.hbase.DoNotRetryIOException:
java.lang.NoClassDefFoundError: org/apache/hadoop/hbase/NoTagsKeyValue
at org.apache.hadoop.hbase.ipc.RpcServer.call(RpcServer.java:2151)
at org.apache.hadoop.hbase.ipc.CallRunner.run(CallRunner.java:101)
at
org.apache.hadoop.hbase.ipc.RpcExecutor.consumerLoop(RpcExecutor.java:130)
at org.apache.hadoop.hbase.ipc.RpcExecutor$1.run(RpcExecutor.java:107)
at java.lang.Thread.run(Thread.java:724)
Caused by: java.lang.NoClassDefFoundError:
org/apache/hadoop/hbase/NoTagsKeyValue
at
org.apache.hadoop.hbase.io.hfile.HFileReaderV2$ScannerV2.formNoTagsKeyValue(HFileReaderV2.java:755)
at
org.apache.hadoop.hbase.io.hfile.HFileReaderV3$ScannerV3.getKeyValue(HFileReaderV3.java:206)
at
org.apache.hadoop.hbase.regionserver.StoreFileScanner.seek(StoreFileScanner.java:161)
at
org.apache.hadoop.hbase.regionserver.StoreFileScanner.backwardSeek(StoreFileScanner.java:478)
NoTagsKeyValue was introduced by:
HBASE-13579 - Avoid isCellTTLExpired() for NO-TAG cases (Ram)
which was marked for 1.1.1
FYI
On Sun, May 24, 2015 at 7:04 AM, [email protected] <
[email protected]> wrote:
> Hi Ted,
>
> Tests are failing for both the branches and master branch as well.
>
> Thanks,
> Rajeshbabu.
> On May 24, 2015 7:21 PM, "Ted Yu" <[email protected]> wrote:
>
> > Did the test fail on branch 4.4-HBase-1.1, branch 4.x-HBase-1.1 or both ?
> >
> > Cheers
> >
> > On Sun, May 24, 2015 at 2:38 AM, [email protected] <
> > [email protected]> wrote:
> >
> > > I ran the tests multiple times locally and they are always passing.
> > >
> > > [INFO] --- maven-site-plugin:3.2:attach-descriptor (attach-descriptor)
> @
> > > phoenix-assembly ---
> > > [INFO]
> > >
> ------------------------------------------------------------------------
> > > [INFO] Reactor Summary:
> > > [INFO]
> > > [INFO] Apache Phoenix ..................................... SUCCESS [
> > > 1.531 s]
> > > [INFO] Phoenix Core ....................................... SUCCESS
> > [49:22
> > > min]
> > > [INFO] Phoenix - Flume .................................... SUCCESS
> > [01:09
> > > min]
> > > [INFO] Phoenix - Pig ...................................... SUCCESS
> > [03:04
> > > min]
> > > [INFO] Phoenix Query Server Client ........................ SUCCESS [
> > > 0.469 s]
> > > [INFO] Phoenix Query Server ............................... SUCCESS
> > [01:47
> > > min]
> > > [INFO] Phoenix - Pherf .................................... SUCCESS
> > [01:05
> > > min]
> > > [INFO] Phoenix - Spark .................................... SUCCESS [
> > > 3.735 s]
> > > [INFO] Phoenix Assembly ................................... SUCCESS [
> > > 21.070 s]
> > > [INFO]
> > >
> ------------------------------------------------------------------------
> > > [INFO] BUILD SUCCESS
> > > [INFO]
> > >
> ------------------------------------------------------------------------
> > > [INFO] Total time: 56:57 min
> > > [INFO] Finished at: 2015-05-24T14:21:52+05:30
> > >
> > > The reason for failures in Jenkings might be HBase 1.1.0 jars are older
> > > ones of it's RCs so getting linkage errors.
> > > Once we remove them from local repository then it will download latest
> > jars
> > > and the tests should pass.
> > > Any one have an idea how to remove the jars form local repository of
> > > jenkins machines?
> > >
> > > Even with -U for mvn it's not downloading the latest jars.
> > > By the time will compare performance with other versions.
> > >
> > > Thanks,
> > > Rajeshbabu.
> > >
> > > On Fri, May 22, 2015 at 1:07 PM, Enis Söztutar <[email protected]>
> wrote:
> > >
> > > > I have pushed PHOENIX-1763 and PHOENIX-1681 to master, 4.4-HBase-1.1,
> > and
> > > > 4.x-HBase-1.1 branches.
> > > >
> > > > I think one unit test is failing (IndexToolIT) which Rajesh is
> taking a
> > > > look.
> > > >
> > > > Enis
> > > >
> > > > On Thu, May 21, 2015 at 6:38 PM, Enis Söztutar <[email protected]>
> > wrote:
> > > >
> > > > > Makes sense.
> > > > >
> > > > > +1 for cutting the RC from the 4.4 release tag. So in your above
> > plan,
> > > I
> > > > > would change step 3 into:
> > > > > (3) Create the 4.4-branch from 4.4.0-HBase-1.0 tag
> > > > >
> > > > > Then we can commit PHOENIX-1763 and PHOENIX-1681 to 4.4-HBase-1.1,
> > > > > 4.x-HBase-1.1 and master.
> > > > >
> > > > > For Phoenix master branch, and 4.5 releases, I think we can drop
> > > > HBase-1.0
> > > > > support if we think that there is a lot of overhead. HBase-1.1 is
> > > rolling
> > > > > upgradable and compatible with 1.0, so users should not be stuck
> with
> > > it
> > > > > for a long time. it is a matter of overhead of keeping one extra
> > branch
> > > > and
> > > > > release vs whether we would want to support Hbase-1.0 users for
> 4.5+.
> > > > Just
> > > > > my 2 cents.
> > > > >
> > > > > Enis
> > > > >
> > > > > On Thu, May 21, 2015 at 5:39 PM, [email protected] <
> > > > > [email protected]> wrote:
> > > > >
> > > > >> Makes sense Nick. Cut the new branch from 4.4.0-HBase-1.0 tag.
> > > > >>
> > > > >> Thanks,
> > > > >> Rajeshbabu.
> > > > >>
> > > > >> On Fri, May 22, 2015 at 5:58 AM, Nick Dimiduk <[email protected]
> >
> > > > wrote:
> > > > >>
> > > > >> > This looks good except for (1) branch should be cut from the
> > > > >> > 4.4.0-HBase-1.0 tag. That way 4.4.0 will be feature-identical
> > across
> > > > all
> > > > >> > releases. Then we'll need to go back over what's landed on
> > > > 4.x-HBase-1.0
> > > > >> > since 4.4.0 was released and bring forward any patches, so that
> > all
> > > 3
> > > > >> > branches are aligned for 4.4.1.
> > > > >> >
> > > > >> > Make sense?
> > > > >> >
> > > > >> > On Thu, May 21, 2015 at 1:35 PM, [email protected] <
> > > > >> > [email protected]> wrote:
> > > > >> >
> > > > >> > > Thanks James. Sure I will run it also.
> > > > >> > >
> > > > >> > > Here is my plan for the release:
> > > > >> > >
> > > > >> > > 1) create the 4.x-HBase-1.1 branch from latest 4.x-HBase-1.0
> > > > >> > > 2) change the hbase version to 1.1.0 and commit PHOENIX-1793
> > > > >> > > 3) Setup jenkins for the branch to run the test suite.
> > > > >> > > 3) cut rc from 4.x-HBase-1.1
> > > > >> > > 4) before vote is going to pass create 4.4-HBase-1.1 branch
> for
> > > > 4.4.x
> > > > >> bug
> > > > >> > > fix releases.
> > > > >> > >
> > > > >> > > I
> > > > >> > >
> > > > >> > > Thanks,
> > > > >> > > Rajeshbabu.
> > > > >> > >
> > > > >> > >
> > > > >> > > On Thu, May 21, 2015 at 9:52 PM, James Taylor <
> > > > [email protected]
> > > > >> >
> > > > >> > > wrote:
> > > > >> > >
> > > > >> > > > Hi Rajeshbabu,
> > > > >> > > > Sounds good. Probably a good idea to run our perf regression
> > > suite
> > > > >> to
> > > > >> > > > get an idea of performance too. I'm fine with a release,
> > though.
> > > > >> > > > Thanks,
> > > > >> > > > James
> > > > >> > > >
> > > > >> > > > On Thu, May 21, 2015 at 6:09 AM, [email protected]
> > > > >> > > > <[email protected]> wrote:
> > > > >> > > > > James,
> > > > >> > > > >
> > > > >> > > > > I have installed HBase 1.1.0 with phoenix-4.4.0-HBase-1.0
> > rc1
> > > +
> > > > >> > > > > PHOENIX-1763 patch at server side. Created
> > > > tables,indexes,uploaded
> > > > >> > data
> > > > >> > > > > with performance script from the new client and performed
> > > > >> > reads/writes
> > > > >> > > > from
> > > > >> > > > > all old clients of versions 4.1.0,4.2.2 and 4.3.1. No
> issues
> > > > >> found.
> > > > >> > > > >
> > > > >> > > > > So I feel we can release with HBase-1.1.0. Running test
> > suite
> > > > and
> > > > >> > seems
> > > > >> > > > > fine.
> > > > >> > > > >
> > > > >> > > > > Thanks,
> > > > >> > > > > Rajeshbabu.
> > > > >> > > > >
> > > > >> > > > >
> > > > >> > > > > On Wed, May 20, 2015 at 7:05 AM, [email protected] <
> > > > >> > > > > [email protected]> wrote:
> > > > >> > > > >
> > > > >> > > > >> James,
> > > > >> > > > >>
> > > > >> > > > >> bq. Will older Phoenix 4.x clients work with the new
> > Phoenix
> > > > 4.4
> > > > >> > > clients
> > > > >> > > > >> against an HBase 1.1 running Phoenix?
> > > > >> > > > >> Mostly they should work. I will quickly check this today.
> > > > >> > > > >>
> > > > >> > > > >> bq. Has Andy's patch already been applied for using
> > > > >> > > > >> the new RS interfaces?
> > > > >> > > > >> The patch at PHOENIX-1763 has required changes after new
> RS
> > > > >> > interface
> > > > >> > > > >> changes.
> > > > >> > > > >>
> > > > >> > > > >> Thanks,
> > > > >> > > > >> Rajeshbabu
> > > > >> > > > >>
> > > > >> > > > >> On Tue, May 19, 2015 at 11:32 PM, James Taylor <
> > > > >> > > [email protected]>
> > > > >> > > > >> wrote:
> > > > >> > > > >>
> > > > >> > > > >>> Will older Phoenix 4.x clients work with the new Phoenix
> > 4.4
> > > > >> > clients
> > > > >> > > > >>> against an HBase 1.1 running Phoenix? If no, we'll need
> to
> > > > >> release
> > > > >> > > > >>> this in a 5.0 version. Has Andy's patch already been
> > applied
> > > > for
> > > > >> > > using
> > > > >> > > > >>> the new RS interfaces?
> > > > >> > > > >>>
> > > > >> > > > >>> Thanks,
> > > > >> > > > >>> James
> > > > >> > > > >>>
> > > > >> > > > >>> On Tue, May 19, 2015 at 10:30 AM, [email protected]
> > > > >> > > > >>> <[email protected]> wrote:
> > > > >> > > > >>> > Hi,
> > > > >> > > > >>> >
> > > > >> > > > >>> > Already HBase-1.1.0 released and we already have patch
> > to
> > > > >> support
> > > > >> > > > with
> > > > >> > > > >>> > it(PHOENIX-1763). I feel we can release 4.4.0 with
> > > > >> HBase-1.1.0 as
> > > > >> > > > well.
> > > > >> > > > >>> >
> > > > >> > > > >>> > What do you guys say?
> > > > >> > > > >>> >
> > > > >> > > > >>> > Thanks,
> > > > >> > > > >>> > Rajeshbabu.
> > > > >> > > > >>>
> > > > >> > > > >>
> > > > >> > > > >>
> > > > >> > > >
> > > > >> > >
> > > > >> >
> > > > >>
> > > > >
> > > > >
> > > >
> > >
> >
>
