[Systems] Google Code-in Tasks (Sysadmin)

[Ignacio Rodríguez]

I would recommend to have only 1 instance, after the task is closed
add it again, remove the VM and create it again.

Maybe have few VM's, one for x task, another for y task, etc.

In that case I think we will have the resources

2015-11-28 21:05 GMT-03:00, Samuel Cantero <scanterog at gmail.com>:
> Hello everyone,
>
> I've been thinking about the way we are going to give access to the VMs and
> where those VMs are going to be hosted.
>
> I guess the appropriate physical machine to host new VMs is freedom. If all
> of you are agree, we must create a template VM for the students who may
> claim one of our tasks which might require a VM. For example: testing
> docker swarm or trying a mailing list tool. How many students
> simultaneously can claim a task? I guess that currently this is
> configurable. Should we have more than one instance of a task? In case we
> have more than one instance, should we have one VM per student? Do we have
> enough resources? What does have to contain the template VM? Do we have
> available IP for the VMs?
>
> Given that we have to give access to a VM, we can add a beginner task such
> as requesting a shell account by sending their 2048 bits public key. This
> public key will be used for a specific account inside the VM with all
> privileges. It would be ideal to prevent sudoers to edit the /etc/sudoers
> file or even to become the root user (get a shell) but I guess that is not
> possible to allow "all commands except ...". There is always a workaround
> to break this barriers. So, what is the best approach to tackle this?
>
> Best regards,
>
> Samuel C.
>
>
> On Fri, Nov 13, 2015 at 8:48 PM, Bernie Innocenti <bernie at codewiz.org>
> wrote:
>
>> I'd be glad to offer guidance for any of these tasks, but can't take an
>> official mentoring role because I'm still supposed to minimize typing
>> even at work.
>>
>> On 11/13/2015 04:12 PM, Martin Abente wrote:
>> > Hey Sam,
>> >
>> > Just send me the email addresses and Ill invite them for you, I already
>> > sent Samuel C. an invite.
>> >
>> > Martin.
>> >
>> > On Fri, Nov 13, 2015 at 6:08 PM, Sam P. <sam at sam.today
>> > <mailto:sam at sam.today>> wrote:
>> >
>> >     Great, I just added the tasks.  I don't know how to add people as
>> >     mentors, I don't see that button.  Maybe Walter is admin and has
>> > it?
>> >
>> >     On Fri, Oct 30, 2015 at 8:58 AM, Sam P. <sam at sam.today
>> >     <mailto:sam at sam.today>> wrote:
>> >
>> >         Hi Samuel,
>> >
>> >         Those tasks look great!  I did not see anything on the gci site
>> >         talking about these kind of tasks.  Maybe we need to ask on irc
>> >         if we get accepted.
>> >
>> >         Also, maybe we could add a task for a sugar3/sugar-toolkit-gtk3
>> >         docs rebuilder?
>> >
>> >         Title:  Sugar3 documentation rebuilder
>> >         Category:  Coding
>> >         Beginner:  No
>> >         Description:
>> >
>> >         In this cycle, we have introduced api documentation for the
>> >         sugar3 api.  This documentation is generated using sphinx.
>> >
>> >         Your task is to create a rebuilder docker container, modeled
>> > off
>> >         our existing rebuilder [1][2].  It must have the correct
>> >         dependencies or mocks installed for the documentation generator
>> >         to run successfully (sphinx imports all the modules, which in
>> >         turn import Gtk, etc.).  The html output should be placed in a
>> >         volume accessible on the host.
>> >
>> >         [1] https://github.com/sugarlabs-infra/help-rebuilder
>> >         [2] https://github.com/sugarlabs-infra/developer-rebuilder
>> >
>> >         Thanks,
>> >         Sam
>> >
>> >
>> >         On Fri, Oct 30, 2015, 4:20 AM Samuel Cantero
>> >         <scanterog at gmail.com <mailto:scanterog at gmail.com>> wrote:
>> >
>> >             Hello guys!
>> >
>> >             Bernie and I have been thinking about some tasks for the
>> > GCI
>> >             contest. As Martin (tch) said in another email, we should
>> >             double check these tasks are "compatible" with the contest
>> >             topics.
>> >
>> >             Below the first draft:
>> >
>> >             *Title*: Analyze and optimize the Sugar Labs wiki
>> performance.
>> >             *Category*: Research / QA / Other?
>> >             *Description*:
>> >             The main documentation for the Sugar Labs organization is
>> >             placed in the Sugar Labs wiki. Speeding up the wiki access
>> >             is important for the community.
>> >             Analyze and optimize our wiki using Google PageSpeed Tools.
>> >             Get a score of 85 or above for the Sugar Labs wiki.
>> >
>> >             Google PageSpeed URL: https://developers.google.com/speed/
>> >
>> >
>> >             *Title*: New mailing list for Sugar Labs.
>> >             *Category*: Research / Other?
>> >             *Description*:
>> >             In Sugar Labs, as most open source organizations, we use
>> >             mailing lists to communicate with one another. Currently,
>> > we
>> >             are using mailman 2.
>> >             Make some research, for example among other open source
>> >             organizations, to propose at least one new open source
>> >             package to deploy a new mailing list for Sugar Labs.
>> >             Implement the chosen software using Docker containers.
>> >
>> >             *Title*: Write a Munin plugin to monitor network resource
>> >             usage on Docker containers
>> >             *Category*: Coding
>> >             *Description*:
>> >             In Sugar Labs, we use Docker container to host important
>> >             services. Monitoring the resource usage is critical for the
>> >             infrastructure health. We have written a Munin plugin to
>> >             monitor CPU and memory usage on Docker containers. The next
>> >             step is  to monitor the network usage on our Docker
>> >             containers. This tasks requires writing a Munin plugin,
>> >             preferably in Perl, for monitoring network resource usage
>> > on
>> >             Docker containers.
>> >
>> >             Docker runtime metrics:
>> >             https://docs.docker.com/articles/runmetrics/
>> >             Git Repo: https://github.com/scanterog/munin-plugin-docker
>> >
>> >             Maybe I have to create a new repo for this on the
>> >             sugarlabs-infra git organization.
>> >
>> >             *Title*: High availability in our Docker infrastructure.
>> >             *Category*: Research / Other?
>> >             *Description*:
>> >             In Sugar Labs, mainly we use Virtual Machines and Docker
>> >             containers to host our main services. We want you to test
>> >             the Docker Swarm functionality in order to create a Docker
>> >             cluster.
>> >
>> >             For this task, you will have two clean Ubuntu VMs in order
>> >             to install and configure the Docker Swarm. We will provide
>> >             you also two Docker images. In order to complete the task,
>> >             the docker cluster and the failover functionality must be
>> >             working.
>> >
>> >             Any advice is welcome.
>> >
>> >             Regards,
>> >
>> >             Samuel C.
>> >             _______________________________________________
>> >             Systems mailing list
>> >             Systems at lists.sugarlabs.org
>> >             http://lists.sugarlabs.org/listinfo/systems
>> >
>> >
>> >
>>
>>
>> --
>>  _ // Bernie Innocenti
>>  \X/  http://codewiz.org
>>
>


-- 
Ignacio Rodríguez
SugarLabs at Facebook
<https://www.facebook.com/pages/SugarLabs/187845102582>