I'm working  on
 * Nicira support for SourceNat, StaticNat and PortForwarding
 * Maven improvements
 * Rpm packaging sample using maven buildsystem

And of course bug finding and fixing whiel we are building our cloudstack cloud 
for our customers.

Cheers,

Hugo

> -----Original Message-----
> From: Kelceydamage@bbits [mailto:kel...@bbits.ca]
> Sent: Saturday, October 27, 2012 7:25 PM
> To: cloudstack-dev@incubator.apache.org
> Cc: cloudstack-dev@incubator.apache.org
> Subject: Re: Moving beyond 4.0.0-incubating
> 
> A python agent is something I would really like. And also is a project I might
> actually be able to work on(Python is my primary language).
> 
> Let me know when you get this going and I will do what I can to assist.
> 
> Sent from my iPhone
> 
> On Oct 27, 2012, at 2:44 AM, Rohit Yadav <rohit.ya...@citrix.com> wrote:
> 
> > Hi,
> >
> > I'm too working on the client based on marvin too
> (https://issues.apache.org/jira/browse/CLOUDSTACK-132), pl. see the
> comments on the issue so we can collaborate on this.
> > Next thing I want to discuss and work on is the agent for KVM.
> >
> > Right now if one install agent on KVM; he had to:
> > - Do it manually on all hosts (not scalable unless some provisioning
> > tool, like chef/puppet is used); required installation of dependencies
> > - Upgrade is manual; properties are stored on host; possibility of
> > loss of these files while upgrade (if upgrade/install scenarios are
> > not very well handled)
> > - Developer development is manual and not smooth; if you change
> > something in agent code, you are required to manually setup agent on
> > hosts
> >
> > Proposed agent:
> > - Mgmt server should ssh and setup the agent; in case of Xen, vmops
> > plugin which is written in Python is deployed and used (much easier as
> > they have xapi service and we're just adding one plugin)
> > - Possibly rewrite agent in Python because:
> >  * Present agent written in Java, comes with its own baggage, required
> openjdk etc.
> >  * Python is preinstalled on most Linux distros
> >  * Writing in Python makes it easy to hack, no hack-build-run loop;
> > just hack-run
> >  * Agent can have it's own environment and run in a virtualenv; the
> > present agent required deps etc. which are not local to the agent, but
> > affects the system in general
> >  * Can reuse libvirt-python (no virsh), example code I'd written in
> > past;
> > https://github.com/bhaisaab/vmcontroller/blob/master/src/vmcontroller.
> > host/vmcontroller/host/controller/hypervisors/LibVirt.py
> > - Rewriting and throwing away code can be a wild ide, so if we decide
> > to not rewrite it, let's discuss some way to automate agent deployment
> > on hosts and make it easier for developers to work with it
> >
> > Regards.
> >
> > ________________________________________
> > From: Sebastien Goasguen [run...@gmail.com]
> > Sent: Saturday, October 27, 2012 2:27 PM
> > To: cloudstack-dev@incubator.apache.org
> > Subject: Re: Moving beyond 4.0.0-incubating
> >
> > Hi,
> >
> > *I emailed with James Martin who has been doing some devcloud
> improvements, I encouraged him to submit his patches even if it's not done.
> >
> > *Prasanna created a marvin-parallel branch to host a threading of the
> integration tests. I will get back to that after the vote.
> >
> > *There is a request for an interactive shell. I just tried jclouds-cli, 
> > it's nice
> but is far from covering all cloudstack can do. So I am starting a prototype
> using the marvin cloudstackAPI.
> >
> > *There is also an important thread on a better AWS support with a newer
> api version.
> >
> > On the more technical front, I think we need to look at network
> > capabilities that are only available with netscaler and provide an
> > open source alternative. Potentially also look at integrating quantum
> > -if you look at it as an open source nicira-
> >
> > -Sebastien
> >
> >
> > On Oct 26, 2012, at 10:07 PM, Edison Su <edison...@citrix.com> wrote:
> >
> >> I am working on the storage subsystem refactor on javelin branch. The
> code I write is under platform/storage directory. I'll send out a detailed
> changes I want to make in the next few days.
> >>
> >>> -----Original Message-----
> >>> From: Chip Childers [mailto:chip.child...@sungard.com]
> >>> Sent: Wednesday, October 24, 2012 2:57 PM
> >>> To: cloudstack-dev@incubator.apache.org
> >>> Subject: Moving beyond 4.0.0-incubating
> >>>
> >>> Hi all,
> >>>
> >>> Obviously we still have to actually pass a vote, and then pass
> >>> another vote within the IPMC, before we can call success with
> >>> 4.0.0-incubating.  (And we may have to do this several more times,
> >>> depending on how things go.)  However, I think it's time to catch up
> >>> on work that's ongoing, ideas for what to do next, and to start to
> >>> plan out what our next community goals are.
> >>>
> >>> Along those lines, perhaps we can all start communicating a little
> >>> better about where other work stands?  I'd like to ask folks that
> >>> are working on features / tools to kick off discussion threads for
> >>> their specific areas.  Questions that would be great to have
> >>> answered
> >>> include:  How's the work going generally?  Are there any design
> >>> decisions that should be talked about?  Any help needed?
> >>>
> >>> Examples that come to mind (in no order, and not exhaustive):
> >>>
> >>> 1- Adding more plugin capabilities to the core design (I assume this
> >>> is what the javelin branch is for)
> >>> 2- RBAC
> >>> 3 - Event handling / message bus integration
> >>> 4 - DevCloud changes
> >>> 5 - Work demonstrated at Citrix Synergy (Cisco product integrations)
> >>> 6 - Auto-scaling
> >>> 7 - Host deployment feature
> >>>
> >>> I'd also like to ask that we start proposing new features to the
> >>> list again (I know that some of this has continued, but I think it
> >>> died down a bit).  It's going to be important to start thinking
> >>> about timing of our next feature release soon (and to be time bound
> >>> about it), so I think we need to ensure that we're all sharing
> >>> current thoughts, status, etc...  And as always, feel free to let me
> >>> know if you disagree!
> >>>
> >>> Thanks all,
> >>>
> >>> -chip
> >

Reply via email to