interesting blog post

[pat eyler]

I don't know if any of you have seen this post about distributed
programming tools, but it seems like an interesting topic for us.

http://www.sriramkrishnan.com/blog/2008/05/what-twitter-and-everyone-else-needs.html

--
thanks,
-pate
-------------------------
Don't judge those who choose to sin differently than you do

http://on-ruby.blogspot.com
http://eldersjournal.blogspot.com

[littleidea]

That's certainly interesting, but I don't see away to do it without
instrumenting each service in the stack individually.

One thing that can help a lot is to find ways to display all the
available data you have with a normalized timeline.

If you have a unified view of deployments (who, what, when) with
traffic (classify what type of and how many requests are coming in)
and the metrics from each server/service, you can reason much more
holistically than you can if you are starting with something like
'things are slow' and trying to work backwards.

There isn't a drop in solution that does this, but there are tools
that can be put together to get you most of the way right now.

On Aug 4, 8:30 am, pat eyler <pat.ey...@gmail.com> wrote:
> I don't know if any of you have seen this post about distributed
> programming tools, but it seems like an interesting topic for us.
>

> http://www.sriramkrishnan.com/blog/2008/05/what-twitter-and-everyone-...