opensand team mailing list archive

[Didier Barvaux <didier.barvaux@xxxxxxxxxxxxxxxxxxxx>]

Andy,

Could you please write all these information on a dedicated wiki? It
will be useful for SOCIS report/evaluation.

I prepared a page for this:
https://opensand.org/support/wiki/doku.php?id=socis_progress_report

Regards,
Didier


Le Mon, 3 Sep 2012 17:57:24 +0200,
Jedrzej Gorski <andy.gorski.dev@xxxxxxxxx> a écrit :

> Hi Dider,
> 
> 
> 
> I have some good news :) My first full test scenario based on 3
> virtual machines is finally working (check attached files - I have
> found some small bug as well)! Many thanks to you for constantly
> updated "installation manual". It helped a lot. Of course in order to
> achieve this task I had to fix some software and hardware issues
> (mainly concerns Oracle Virtual Box which I describe latter) and lack
> of RAM in my lab equipment (2 GB is not enough for basic virtual test
> infrastructure - SAT, ST, GW; I recommend >= 3GB of RAM). I still
> have some issues with source compilation but I'll describe it after
> my experience with packages.
> 
> 
> 
> 1. Basic testing for package compilation
> 
> 
> Before I describe issues associated with my OpenSAND latest
> installation I need to catch up with some other topics. According to
> my previous email - with attached photo of working manager  - to
> perform my first successful compilation I use only packages. I create
> a single virtual machine (Oracle Virtual Box) on a fresh hardware
> ("Main VM" on a desktop with Intel Pentium Dual CPU 1,81 GHz platform
> with 2GB of RAM) based on ubuntu 10.04 LTS as a guest and Windows XP
> as a host system. That was successful test with one daemon (SAT) and
> sand manager (check OpenSAND - first sucess.jpg and log file).
> 
> 
> 
> I've conducted similar test scenario on my DELL laptop (Intel Core
> Duo 1,87 GHz with 2GB of RAM) with the same software configuration
> but I still have Python related problems which I describe in previous
> post. In reply to your suggestion I've send results of tests which
> you’ve asked me to do. I think I mass up to much with that ubuntu
> VM :) I based this opinion on a fact that I perform several test with
> packages compilation (based on updated manual) and I couldn't
> reproduce that errors on the other hardware. I've put the result of
> that try-out as a reply to your suggestions from one of previous
> posts -> Python issue during Open Sand platform installation (3.09).
> The same post covers issue related to "missing pid" which can be
> found in attached log file - first section.
> 
> 
> 
> 2. Extended testing for package compilation
> 
> 
> 
> Because of that small success (first successful compilation with SAT
> daemon and sand manager from last week) I was full of hopes for a
> first full network test. Unfortunately it wasn't as easy as I
> expected. I had quite a lot of problems with 3 VM's working together
> under Virtual Box ("Main VM" and two clones of that virtual machine
> on a desktop with Intel Pentium Dual CPU 1,81 GHz platform with 4GB
> of RAM). Finally I have manage to create full test scenario
> infrastructure on my "strongest" hardware configuration (check
> OpenSAND - sucess 3.jpg). At the moment I have two basic test
> platforms and I'm working on Virtual Box configuration (and in a free
> time I'm testing VM Ware team configuration). Test bed configuration
> has one computer with enough resources to handle at least 3 virtual
> machines and host system - Windows XP.
> 
> 
> Some issues with Virtual Box are described in separate post -> Test
> infrastructure for OpenSAND - Issues with Oracle Virtual Box
> 
> 
> 3. New hardware arrived
> 
> 
> 
> Since Monday (27.08) I have two additional machines with AMD Athlon X2
> processors (64 bits) with 4GB of RAM each. At the moment I'm using
> Windows XP 32 bit as a host system just to have some simple and
> reliable host. I have install 3 independent VM's just to have
> something independent and just in case is not replicating any issues
> which might show up in "Main VM". Above that because of this new
> hardware I'm able to do some tests with 64-bits operating systems -
> possible VM's with windows 7 or WInXP 64-bit edition hosts or some 64
> bit linux editions. I'm open for suggestions :) At the moment I'm
> using as simple configuration as possible - My main goal is to reach
> a point where source code is compiling as easy as packages.
> 
> 4. Operational issue with OpenSAND manager
> 
> Through attached log file one can have an impression how this test
> infrastructure works now. In the section generated through manager
> you can find an information that manager has detected duplicated
> devices. In this case manager is working on the same VM as gateway
> daemon. This scenario was generated on single VM without presence of
> the other elements of test infrastructure.
> 
> powerxp@powerxp-desktop:~/opensand-0.9.0-deb$ sand-manager
> 2012-08-31 11:32:37,446 - sand-manager - INFO - Welcome to OpenSAND
> Manager !
> 2012-08-31 11:32:37,446 - sand-manager - INFO - Initializing platform,
> please wait...
> 2012-08-31 11:32:37,791 - sand-manager - INFO - Find gw at address
> 10.0.4.15 2012-08-31 11:32:37,842 - sand-manager - INFO - Find gw at
> address 192.168.19.2
> 2012-08-31 11:32:37,843 - sand-manager - WARNING - gw : duplicated
> service received at address 192.168.19.2
> 2012-08-31 11:32:37,844 - sand-manager - INFO - Find gw at address
> 192.168.18.3
> 2012-08-31 11:32:37,874 - sand-manager - WARNING - gw : duplicated
> service received at address 192.168.18.3
> 
> That would be all topics which I wish to report in this post.
> 
> I'm looking forward for your feedback.
> 
> Greetings,
> 
> Andy