Questions for a white paper involving a demo install of Typo


#1

All,

I’ve been asked to put together a white paper about Ruby on Rails.
Rather write yet another article with a simple proof-of-concept
application tutorial, I thought it might be more interesting to talk
about scaling issues. With that in mind, my current plan is to set up
Typo in a multiple server environment, turn off caching, and start load
testing it to see how many FCGI listeners on each box gets the best
performance. The results would form the basis of the paper.

My hope is that an article specifically about how Rails apps get
deployed in high-volume, high-availability environments would be more
interesting than another round of “now let’s do script/generate
scaffold…”

I’d appreciate any feedback you might have to offer, both on whether or
not you’d find such a paper interesting, and on any particular
implementation issues that spring to mind. Maybe it’ll even be useful
if you happen to have a Typo-driven blog so popular that you need a
half-dozen dedicated servers just to run it. :wink:

Many thanks for your time,

Sam L. removed_email_address@domain.invalid
Comm Solution Development
Unisys
phone: 610.648.3702


#2

One suggestion: turn off all of the sidebars. The current
sidebarimplementation has some performance issues with Typo with at
leastsome versions of rails; I haven’t benchmarked it recently.

Scott
On 4/11/06, Leibowitz, Samuel J removed_email_address@domain.invalid wrote:>
All,>> I’ve been asked to put together a white paper about Ruby on
Rails.> Rather write yet another article with a simple proof-of-concept>
application tutorial, I thought it might be more interesting to talk>
about scaling issues. With that in mind, my current plan is to set up>
Typo in a multiple server environment, turn off caching, and start load>
testing it to see how many FCGI listeners on each box gets the best>
performance. The results would form the basis of the paper.>> My hope is
that an article specifically about how Rails apps get> deployed in
high-volume, high-availability environments would be more> interesting
than another round of “now let’s do script/generate> scaffold…”>> I’d
appreciate any feedback you might have to offer, both on whether or> not
you’d find such a paper interesting, and on any particular>
implementation issues that spring to mind. Maybe it’ll even be useful>
if you !
happen to have a Typo-driven blog so popular that you need a> half-dozen
dedicated servers just to run it. ;)>> Many thanks for your time,>> Sam
Leibowitz removed_email_address@domain.invalid> Comm Solution Development>
Unisys> phone: 610.648.3702>>
_______________________________________________> Typo-list mailing list>
removed_email_address@domain.invalid>
http://rubyforge.org/mailman/listinfo/typo-list>


#3

Oh, and one other thing–if you use the current Typo trunk and
runmemcached on your servers, then you should be able to use
Rails’memcache fragment cache implementation, so you’ll have a working
cachespread across multiple front-end servers with correct
cacheinvalidation.

Scott
On 4/11/06, Scott L. removed_email_address@domain.invalid wrote:> One suggestion: turn
off all of the sidebars. The current sidebar> implementation has some
performance issues with Typo with at least> some versions of rails; I
haven’t benchmarked it recently.>>> Scott>> On 4/11/06, Leibowitz,
Samuel J removed_email_address@domain.invalid wrote:> > All,> >> > I’ve been
asked to put together a white paper about Ruby on Rails.> > Rather write
yet another article with a simple proof-of-concept> > application
tutorial, I thought it might be more interesting to talk> > about
scaling issues. With that in mind, my current plan is to set up> > Typo
in a multiple server environment, turn off caching, and start load> >
testing it to see how many FCGI listeners on each box gets the best> >
performance. The results would form the basis of the paper.> >> > My
hope is that an article specifically about how Rails apps get> >
deployed in high-volume, high-availability environments would be more> >
interesting !
than another round of “now let’s do script/generate> > scaffold…”> >>

I’d appreciate any feedback you might have to offer, both on whether
or> > not you’d find such a paper interesting, and on any particular> >
implementation issues that spring to mind. Maybe it’ll even be useful>
if you happen to have a Typo-driven blog so popular that you need a> >
half-dozen dedicated servers just to run it. ;)> >> > Many thanks for
your time,> >> > Sam L. removed_email_address@domain.invalid> > Comm
Solution Development> > Unisys> > phone: 610.648.3702> >> >
_______________________________________________> > Typo-list mailing
list> > removed_email_address@domain.invalid> >
http://rubyforge.org/mailman/listinfo/typo-list> >>


#4

“Scott L.” removed_email_address@domain.invalid writes:

One suggestion: turn off all of the sidebars. The current
sidebarimplementation has some performance issues with Typo with at
leastsome versions of rails; I haven’t benchmarked it recently.

I believe that Rails 1.1 is supposed to have improved component
performance. Which is nice.