greetings with a proposal
by Amir Sanjar
Hi all,
For the last two years my team at IBM has been involve in maintaining
Apache hadoop and related
projects ( i.e. HBase, HIVE, PIG,..). We have extensive knowledge in area
of porting Apache hadoop projects
to other platforms (i.e. IBM JAVA, POWER, SystemZ,..). We would like to
bring this knowledge to aid expanding fedora's hadoop
ecosystems.
To start, I would like to propose adding support for following platforms
(Sorted by importance) to fedora hadoop:
1) IBM POWER
2) IBM JAVA 7 (post October release) - This might be belong to Fedora JAVA
SIG
3) IBM SystemZ
Best Regards
Amir Sanjar
IBM PowerLinux Hadoop projects Architect and lead
IBM Senior Software Engineer
Phone# 512-286-8393
Fax# 512-838-8858
7 years, 4 months
Hive packaging wiki
by Peter MacKinnon
Hello friends,
I have assembled a Hive packaging wiki
<https://fedoraproject.org/wiki/Hive_packaging>. Due to the nature of
the Hive ivy build and the way it can compile to several versions of
Hadoop, it is still a little open with regards to missing dependencies.
However, the next certain one will be the datanucleus deps which I will
be working on. This dep will need some hacking to eliminate some source
level dependencies. That project is set up to build all supported RDBMS
adapters, but it includes an Oracle JDBC dep which won't make it into
Fedora obviously.
A repo <https://github.com/fedora-bigdata> will be set up for Hive
itself soon once the datanucleus package has sufficiently progressed.
Let me know if you have any questions.
\Pete
--
Peter MacKinnon
MRG Grid/Big Data
Red Hat Inc.
Raleigh, NC
7 years, 4 months
New hadoop package (2.0.5-9)
by Robert Rati
There is a new hadoop package that has been pushed to F20 testing. This
new package includes new client and test subpackages that help help
resolve hadoop related dependencies for parts of the ecosystem. It also
fixes a few issues with the lichdfs package and a file conflict with
qt-devel.
Rob
7 years, 4 months
Fwd: [PROPOSAL] Aurora for Apache Incubation
by Tim St. Clair
I'm game, but my bandwidth will be limited.
https://wiki.apache.org/incubator/AuroraProposal
I was playing in this space in condor for a while, some of the isolation work (auto-fuse-mounting,chroots,cgroups,namespaces) will likely translate.
Cheers,
Tim
----- Forwarded Message -----
> From: "Dave Lester" <dave(a)ischool.berkeley.edu>
> To: dev(a)mesos.apache.org
> Sent: Wednesday, September 4, 2013 8:00:13 PM
> Subject: Re: [PROPOSAL] Aurora for Apache Incubation
>
> The good news is that we seem to be just weeks away from being able to
> initially share the code publicly. The purpose of sharing the proposal to
> the incubator was to gain interest in the project, and recruit Apache
> Incubator mentors.
>
> If you have a specific need for Aurora right away and wish to become an
> initial Apache contributor, there is a possibility that we could reach an
> agreement to provide an early release. (Primary concern here is IP as we
> continue to scrub the code) If you're interested, contact me directly. For
> the vast majority of folks on this list, I'd suggest you hang in there and
> we'll make sure to announce any releases on the Mesos dev and user lists.
>
> Best,
> Dave
>
> On Sat, Aug 31, 2013 at 2:17 PM, vonPuh fonPuhendorf <
> vonpuhfonpuhendorf(a)gmail.com> wrote:
>
> > What means by request? I would like to download it.
> > На 27.08.2013 20:58 "Dave Lester" <dave(a)ischool.berkeley.edu> написа:
> >
> > > Hi Jim,
> > >
> > > The code has not yet been released, but may be made available upon
> > request
> > > as we work toward open sourcing it.
> > >
> > > Best,
> > > Dave
> > >
> > > On Tue, Aug 27, 2013 at 8:55 AM, Jim Morrison <jmorrison(a)quantcast.com
> > > >wrote:
> > >
> > > > Hi Dave,
> > > >
> > > > Is the code released anywhere?
> > > >
> > > > Jim Morrison
> > > > jmorrison(a)quantcast.com
> > > >
> > > >
> > > >
> > > > On Aug 26, 2013, at 3:39 PM, Dave Lester <dave(a)ischool.berkeley.edu>
> > > > wrote:
> > > >
> > > > > Hi All,
> > > > >
> > > > > We're pleased to share a draft ASF incubation proposal for Aurora, a
> > > > > service scheduler used to schedule jobs onto Apache Mesos that we've
> > > > > developed at Twitter. Aurora provides all of the primitives necessary
> > > to
> > > > > quickly deploy and scale stateless and fault tolerant services in a
> > > > > datacenter. The complete proposal can be found online:
> > > > > https://wiki.apache.org/incubator/AuroraProposal
> > > > >
> > > > > In particular, we'd love to add additional mentors to the project.
> > Your
> > > > > feedback is appreciated, which you can make on the
> > > > > general(a)incubator.apache.org mailing list.
> > > > >
> > > > > Thanks!
> > > > >
> > > > > Dave
> > > >
> > >
> >
>
--
Cheers,
Tim
7 years, 4 months
Fwd: [Test-Announce] Call for Fedora 20 Test Days
by Matthew Farrellee
Is anyone interested in doing one of these?
The Hadoop package seems like a good candidate.
Best,
matt
-------- Original Message --------
Subject: [Test-Announce] Call for Fedora 20 Test Days
Date: Mon, 26 Aug 2013 22:16:13 -0600
From: Mike R <roshi(a)fedoraproject.org>
Reply-To: Development discussions related to Fedora
<devel(a)lists.fedoraproject.org>
To: devel(a)lists.fedoraproject.org, test(a)lists.fedoraproject.org
Hey all, it's a bit late but it's time to consider putting in your
requests for Test Days for Fedora 20.
As a recap, a Test Day is an online event aimed at testing a specific
feature of an upcoming Fedora Release. By utilizing IRC for
organization and coordination and a Wiki page for instructions and
results Test Days are easy to organize. Anyone can request to host a
Test Day or request that the QA team help you out with the
organization of the test day. A test day can be ran for any feature or
area of a distribution that focused testing would be useful for. More
information on test days can be found here:
https://fedoraproject.org/wiki/QA/Test_Days .
To propose a Test Day, file a ticket on the QA Trac. A full
explanation can be found here:
https://fedoraproject.org/wiki/QA/Test_Days/Create . The SOP for
hosting a test day is here:
https://fedoraproject.org/wiki/QA/SOP_Test_Day_management . Typically
Test Days are on a Thursday, or for a series of Test Days over a
Tues-Thursday period (this is how we usually run the X Test Week, for
instance).
There are many days available starting today 2013-08-26 all the way
through 2013-11-26. Take into consideration the release schedule here:
http://fedoraproject.org/wiki/Releases/20/Schedule . It might be a
good idea to schedule your test day before the Final Change Deadline.
And finally, if you really want to run a Test Day in a specific
time-frame due to the development schedule, but the Thursday slot for
that week is full, we can add a slot on another day. We're flexible!
Just put in your ticket the date or time-frame you'd like, and we'll
figure it out from there.
If you have any questions about Test Days or the process, please don't
hesitate to contact me or any other QA Team member in #fedora-qa on
Freenode.
Thanks and happy testing!
Mike Ruckman
Fedora QA
IRC: roshi
--
devel mailing list
devel(a)lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
7 years, 4 months
Re: [distcomp] Need hadoop-common packaged for ecosystem
by Peter MacKinnon
Nevermind, I see the problem. I need to "enhance" my local Ivy resolver
to check /usr/lib also.
\Pete
On 09/04/2013 04:10 PM, Peter MacKinnon wrote:
> Hey Rob, I think we're going to need to package the hadoop-common jar
> and frags in the next spin. Hive (and likely others) will need it.
> I can take it on if you like.
>
> \Pete
>
> [ivy:resolve]
> [ivy:resolve] :: problems summary ::
> [ivy:resolve] :::: WARNINGS
> [ivy:resolve] module not found:
> hadoop#hadoop-common;2.0.5-alpha
> [ivy:resolve] ==== fedora1: tried
> [ivy:resolve] -- artifact
> hadoop#hadoop-common;2.0.5-alpha!hadoop-common.jar:
> [ivy:resolve] /usr/share/java/hadoop-common.jar
> [ivy:resolve] ==== fedora2: tried
> [ivy:resolve] -- artifact
> hadoop#hadoop-common;2.0.5-alpha!hadoop-common.jar:
> [ivy:resolve] /usr/share/java/hadoop/hadoop-common.jar
> [ivy:resolve] ::::::::::::::::::::::::::::::::::::::::::::::
> [ivy:resolve] :: UNRESOLVED DEPENDENCIES ::
> [ivy:resolve] ::::::::::::::::::::::::::::::::::::::::::::::
> [ivy:resolve] :: hadoop#hadoop-common;2.0.5-alpha: not found
> [ivy:resolve] ::::::::::::::::::::::::::::::::::::::::::::::
>
>
--
Peter MacKinnon
MRG Grid/Big Data
Red Hat Inc.
Raleigh, NC
7 years, 4 months
Fwd: Broken dependencies: glusterfs-hadoop
by Matthew Farrellee
What was the fix for this?
-------- Original Message --------
Subject: Broken dependencies: glusterfs-hadoop
Date: Mon, 2 Sep 2013 10:56:29 +0000 (UTC)
From: buildsys(a)fedoraproject.org
To: glusterfs-hadoop-owner(a)fedoraproject.org
glusterfs-hadoop has broken dependencies in the F-20 tree:
On armhfp:
glusterfs-hadoop-2.1.2-1.fc20.noarch requires
mvn(org.apache.hadoop:hadoop-common)
Please resolve this as soon as possible.
7 years, 4 months