On Monday 27 June 2005 06:23 pm, Josh Hunholz wrote:
> Wilkins, Vern wrote:
> >What would the web interface do, or more importantly, what would it do
> > that webmin doesn't do already?  I'm not a programmer, but it would seem
> > that writing a module for webmin to handle something like portage
> > administration would be much easier, and probably more quickly result in
> > a more complete administration interface for a gentoo machine.  I'm not
> > doubting that one could develop a better web interface or tool than
> > webmin, for Gentoo, just curious.
>
> The idea behind this is far more than Webmin does. One of the coolest
> features looks to be a way to keep multiple Gentoo boxes up to date with
> security fixes (so it will tell you what boxes have what that need
> updating, etc.) and also more of a large-scale administration of
> multiple servers from one interface (where Webmin does one server).
>
> --Josh Hunholz
>

Or, an expansion on the whole idea may come as something like this:
A cross-platform distribution (based on Gentoo ;) ) with the ability to 
cluster (however it be necessary).  Comes with tools (written in python?) for 
image creation, whereas images are "stacks" of packages prebuilt/compiled by 
the administrator.  You might specify the stack or meta-ebuild as a set of 
compiled packages.  Once an image is created it functions as an enterprise 
component.  The image then is run under a virtual machine that supports 
running on a cluster.  There are sets of tools to manage the virtual 
machines, and allow for primary and backup virtual services to be dynamically 
loaded and configured, or shutdown when not needed.  

This creates a dynamic network where you would design it with statefulness in 
mind.  When I say statefulness, I am referring to how state changes in a 
network.  One minute things are fine, the next, there are floods of traffic 
pumping through your email server and ldap is being floored because you 
didn't have time to replicate it into a secondary.  With that in mind, we 
specify something as responding within a time limit as maintaining a state.  
The moment it breaks that limit, action is taken to recover the state.

Network operation can be dynamically altered by setting up (virtual) bridges 
between the "enterprise components" and dynamically routing traffic between 
each of the components.  Each enterprise component would sit as a group of 
almost duplicate virtual machines dedicated to managing one task.  When load 
goes up, more virtual machines are powered up.  When load goes down, some are 
shutdown.  We could even go as far as incorporating AI to preload these to 
match trends.

Some advantages may be:
* Redundancy - something goes down, the service can be replicated as more 
virtual machines are powered on.
* Load balance all services as one - Each service is balanced (application 
specific, or network traffic) across multiple virtual hosts, but sits on top 
of the cluster.  The cluster distributes the load of the virtual machines.
* Hardware failure - Hardware fails, nothing goes down?  I like that :)
* Statefulness - all that needs storage is the states of the machines, and the 
data that circulates, I'm thinking storage area network... (i.e. providing 
the state, and a data source, the cluster builds itself to maintain the 
state, then provides services to circulate the data)
* Cost - finally a use for all that old equipment we have laying around

This may not be the next step, but the next step might lead to this.

Robert Larson
-- 
[email protected] mailing list

Reply via email to