there are many things u can do about this which are a lot scalable and
more longterm solutions than moving to a singapore data center.
client side caching by setting the http cache-control header.
server side page / fragment / query caching in memcached.
THEN
reverse proxy caching with the proxy placed into singapore.
it requires more architecting though, of course.
--
tom
if your customer base is mainly global, the US is still best connected.
if you serve a lot of customers you can use CDN Networks or global load balancers.
AWS and SoftLayer are good providers. Also your local Webhoster if you like to serve a specific market.
br
> --
> You received this message because you are subscribed to the Google Groups "Singapore Ruby Brigade" group.
> To post to this group, send email to singap...@googlegroups.com.
> To unsubscribe from this group, send email to singapore-rb...@googlegroups.com.
they might change their minds about safety and stuff, especially if u
explicitly bill them for system administration and explain what can
happen if it's not done.
let them know what is the cost equivalent of such a reliable and
redundant system which u can get at heroku.
(calculate with 1 voxel server + 1 hot standby / backup ec2 instance +
their maintenance cost to keep them up to date and synced...)
at the end it will look really expensive, but they might say:
"okay, but we can give up on reliability, just get 1 voxel server"
let's say it halves the cost.
that cost u save is actually the cost of having a less reliable, more
time sucking system locally.
plus the other half of the cost is still a magnitude higher than heroku's cost.
personally, i've experienced the bad effect of having huge performance
/ storage capacity at your disposal.
u start to do programming in a very lousy way, which wastes these
resources at a lot higher rate.
it means u will out grow the dedicated hardware fast and there is
nowhere to go afterwards, because adding more machines, together with
the higher maintenance cost just doesnt worth it.
at the end u should sit down for a month or two and refactor and
re-architect your solution.
which is not cheap either and anything, but fun, plus it's a quite
risky and complicated continuous migration...
--
tom