Opened 10 years ago

Closed 10 years ago

#186 closed (fixed)

Integrate DRAGON with a GENI CF

Reported by: Christopher Small Owned by: chris@maxgigapop.net
Priority: major Milestone: MAX: Integrate DRAGON/GENI control framework
Component: MAX Version: SPIRAL1
Keywords: Cc: Christopher Small
Dependencies:

Description

Design and implement code to integrate DRAGON with a GENI control framework and provide documentation of the integration

Change History (2)

comment:1 Changed 10 years ago by Christopher Small

Work is underway


Subject: Re: Open GENI MAX milestones Date: Tue, 18 Aug 2009 19:32:36 -0400 From: Chris Tracy <chris@maxgigapop.net> To: Christopher Small <csmall@bbn.com> CC: Peter O'Neil <poneil@maxgigapop.net> References: <4A89A725.5010109@bbn.com>

Hi Chris,

[...] The only piece I'd like to see before checking them off is instructions for use (e.g. create the "Accessing Slices" page linked to at the bottom of http://geni.maxgigapop.net/twiki/bin/view/GENI/RequestAccess). My guess is that it's only a few minutes work to write that page; let me know when it's up and I'll check the tasks off and tell Chip and Henry.

Yes -- I am working on these pages now, I am also anxious to get these posted for the users who are in the process of setting up their accounts. There is actually quite a lot of content going into these, mostly because nobody seems to have documented any good examples of using the Princeton GENIWrapper. The closest thing I have seen is http://svn.planet-lab.org/svn/geniwrapper/branches/0.8/docs/geniwrapper.pdf, which describes the PlanetLab implementation of the SFA -- particularly the section at the end that documents the command-line usage of sfi. The examples given in that document only give a very basic introduction as to how to use GENIWrapper to interact with a slice.

comment:2 Changed 10 years ago by Christopher Small

Resolution: fixed
Status: newclosed

Work is completed, modulo bugs in the PlanetLab GENI code. Work-around in place for now (hand-creation of slices for experimenters).


Subject: Re: [GENI: geni] #186: Integrate DRAGON with a GENI CF Date: Sat, 22 Aug 2009 02:27:50 -0400 From: Chris Tracy <chris@maxgigapop.net> To: Christopher Small <csmall@bbn.com> CC: Peter O'Neil <poneil@maxgigapop.net> References: <046.f7f230f91fbe22631dab7887dc2f61e0@geni.net> <055.7fa93ec5c99a305eeba905d647749699@geni.net>

Hi Chris,

I posted a message to the PlanetLab devel list a bit earlier this evening, bcc'ing you and Peter.

I have been attempting to get the Princeton SFA code (what was formerly known as the GENIWrapper code) to create slices on our PlanetLab deployment.

The goal is to allow any user who has signed up for an account to create their own slices as they need them. The Princeton SFA code is supposed to do this, but it does not seem to be quite there yet..

For now, I am resorting to creating the slices myself until the bugs are worked out of the SFA code. The 6 new users who now have accounts (in addition to 6 users from Princeton) have all been granted slices manually. The primary developer who I have been working with (Sapan Bhatia) has been on vacation and will be returning on Monday 08/24. I am confident that once he returns, we can work through the remaining issues that are preventing users from creating their own slices.

In any case, the 'Accessing Slices' page has been completed and contains information about how to install the SFA software to gather information about your slice, how to login to nodes within your slice, and how to configure/use VLAN subinterfaces which terminate on PlanetLab nodes.

I am also planning to e-mail a link to this page to all of the current users who have setup their accounts in the past week.

Thanks! -Chris

Note: See TracTickets for help on using tickets.