Cloudbees RUN@Cloud

17 views
Skip to first unread message

Steven Herod

unread,
Jan 31, 2011, 7:14:39 PM1/31/11
to The Java Posse
Looks interesting, seems more in line with VMForce and Amazon Elastic
Beanstalk than GAE

http://www.cloudbees.com/run.cb

Wonder if all these slightly rounder offering will cause changes in
GAE?

Christian Catchpole

unread,
Jan 31, 2011, 7:29:46 PM1/31/11
to The Java Posse
Looks awesome. But I still think those who want changes to GAE are in
denial about what it takes to scale.

Josh McDonald

unread,
Jan 31, 2011, 7:47:51 PM1/31/11
to java...@googlegroups.com
Isn't that the entire point of platform-as-a-service, tho? Let somebody else worry about what it takes to scale.

--
You received this message because you are subscribed to the Google Groups "The Java Posse" group.
To post to this group, send email to java...@googlegroups.com.
To unsubscribe from this group, send email to javaposse+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.



Christian Catchpole

unread,
Feb 1, 2011, 6:09:19 PM2/1/11
to The Java Posse
Exactly, but if Google let you create Threads, run long requests etc,
then people would be writing all sorts of crazy stuff and blaming
google when it either performed badly and didn't actually remain in
sync as the number of instances grew.

From what I see of cloudbees they are trying to offer the best of
both: eg. you can create files but they can be destroyed at any moment
etc.

It looks like Google looked at why PHP, Perl etc work as web delivery
languages. Not because of the language, but because they are
stateless. You can have a million PHP files on your server and never
need to worry about how much memory you have allocated each "app".
They get compiled, run, and thrown away - or cached now?. is it better
now? I don't know and I guess the point is it doesn't matter.

If people want to add inherently singleton functionality to a cloud
array, then I guess it's their fault if they get it wrong. Google
look like they don't want to give people that much rope.

Perhaps part of the problem is that people are trying to mix tiers.
The presentation tier shouldn't be doing the cron jobs. Google have
expanded cron jobs to 10 mins, which still might not be enough for
some people.

Perhaps the solution is something like separate tier types. Have
"back-end" instances which can do a lot more than the scalable front
tiers but there are fewer of them.


On Feb 1, 10:47 am, Josh McDonald <josh.g.mcdon...@hotmail.com> wrote:
> Isn't that the entire point of platform-as-a-service, tho? Let somebody else
> worry about what it takes to scale.
>
> On 1 February 2011 10:29, Christian Catchpole <ato...@catchpole.net> wrote:
>
>
>
> > Looks awesome.  But I still think those who want changes to GAE are in
> > denial about what it takes to scale.
>
> > On Feb 1, 10:14 am, Steven Herod <steven.he...@gmail.com> wrote:
> > > Looks interesting, seems more in line with VMForce and Amazon Elastic
> > > Beanstalk than GAE
>
> > >http://www.cloudbees.com/run.cb
>
> > > Wonder if all these slightly rounder offering will cause changes in
> > > GAE?
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "The Java Posse" group.
> > To post to this group, send email to java...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > javaposse+...@googlegroups.com<javaposse%2Bunsubscribe@googlegroups .com>
> > .
Reply all
Reply to author
Forward
0 new messages