Tanja Ertl wrote:
Hi Emmanuel,
My client fails already with 75 requests, I don't think that this is
overwhelmed compared to 900 requests from 9 different machines.
How did your client setup look like? If I introduce a delay of 5 ms between the
Threads startup, all seems to work fine. The problem seems to be that the
request arrive all exactly at the same time.
Then it makes sense. The server have a backlog of 50 connections, so if
all your 75 clients are sending requests exactly at the same time, then
you will get some failures. However, this is not exactly a real life
example.
I would suggest that you introduce a ramp up period, where each client
connects every 100 ms, for instance.
If you could create a JIRA for this problem, we can consider adding a
configurable backlog in the server.
--
--
cordialement, regards,
Emmanuel Lécharny
www.iktek.com
directory.apache.org
