How to count dropped connections

[shimi]

On Tue, Jun 9, 2009 at 1:44 PM, Amos Shapira <amos.shapira at gmail.com> wrote:

> Hello,
>
> We are running a cluster of web servers of various functions on top of
> CentOS 5, and would like to know whether and how many requests we
> might be dropping because of insufficient provisioning. The cluster
> could be hit by a a few millions requests per day soon and we'd like
> the stats to be there both now (in case we miss something obvious) and
> when the crowd comes charging.
>
> Googl'ing around for explanations of "netstat -s" and the files under
> /proc/net didn't come up with anything which looks like the definite
> answer, except maybe "packets pruned from receive queue because of
> socket buffer overrun", but I'm not sure what it counts.
>
> Does anyone know how to get the stats I want?
>


At what layer do you define "dropping a request" ? Not accepting a TCP
connection (4) ? Failure to  complete the request from the reverse proxy to
the backend servers (HTTP error) (assuming you have backend servers - the
network structure is not obvious from your original message)?

What answers the TCP requests to port 80?

Do you use efficient HTTP handlers already, e.g. Lighttpd or even better,
nginx? :)

-- Shimi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.huji.ac.il/pipermail/linux-il/attachments/20090609/be7afff6/attachment.html>