[sakai2-tcc] Infrastructure discussion for future meeting?

John Bush john.bush at rsmart.com
Wed Mar 13 11:25:00 PDT 2013


Right, I get that.  I guess I'm a little torn here.  I think David
last comments are on mark, to really get Sakai to the next level in
terms of scale and reliability we need to start componentizing things.
 There's a point when moving towards that makes the architecture much
more complicated and catering to the small guy becomes harder.  I
think if possible I would to prefer one solution that scales, so that
you don't need to switch your whole architecture when you hit some
point, that is what I think for example ElasticSearch brought to the
table regarding search, it works for the small guy and the big guy.
There is giant loss of focus that we have historically suffered from
making things work in 18 different ways.  I mean everyone that runs on
mysql and has to figure out how to get Oracle up and running knows
what I'm talking about. Its a giant pain, and often things get done
half done because of it.

I think I have enough to churn on here for awhile, I didn't mean to
highjack this thread, its just this is one thing I've been focused on
for awhile, and I want to see it move forward.  I think I'll spend
some time thinking and then come back when maybe I have some ideas
that will hit everyone's needs and maybe try to articulate the problem
space a bit more.

On Wed, Mar 13, 2013 at 11:09 AM, Aaron Zeckoski <azeckoski at unicon.net> wrote:
> Mostly I mean running in the same tomcat (think the Sakai demo). If it
> won't work as part of the Sakai binary/demo then that's a non-starter
> IMO.
> -AZ
>
>
> On Wed, Mar 13, 2013 at 2:04 PM, David Adams <da1 at vt.edu> wrote:
>> Aaron Zeckoski wrote:
>>> Personally, I would except that any major functionality would work (1)
>>> on a basic single server installation and also would work (2) on a
>>> multiple tomcats installations and finally could optionally work (3)
>>> with a separate server (for the really big installations). I don't
>>> think options that don't support at least 1 and 2 are suitable for
>>> core Sakai (as a contrib thing, sure).
>>
>> This makes sense. But by "single server" do you mean "single process"?
>> If the MQ server could run in a simple pre-defined config on localhost
>> is that too far to tread?
>>
>> -dave
>> _______________________________________________
>> sakai2-tcc mailing list
>> sakai2-tcc at collab.sakaiproject.org
>> http://collab.sakaiproject.org/mailman/listinfo/sakai2-tcc
>
>
>
> --
> Aaron Zeckoski - Software Architect - http://tinyurl.com/azprofile



--
John Bush
602-490-0470


More information about the sakai2-tcc mailing list