On Fre 04.05.2007 11:36, Robert Iakobashvili wrote:
>On 5/4/07, Aleksandar Lazic <al-curllibrary_at_none.at> wrote:
>>
>>On Mit 02.05.2007 20:49, Robert Iakobashvili wrote:
>> >
>> >A server side able to run 50-100 K connections may be lighttpd
>> >(epoll mode with fd-max increased) with another piece of a good HW,
>> >where a single CPU server with 2GB memory might be enough.
>>
>>wow that's a lot.
>>
>>How many clients have you used against the lighttpd server?!
>
>The above is the description of HW, that we need to reach, hopefully,
>the 50-100 K number of loading clients. Still, other bottlenecks may be
>discovered during the planned experiments, but this is the goad to
>reach it in several months.

Ah ok.

If it is not to disrespectful and arrogating have you the possibility to
make one or two tests with nginx as replacement for lighty, if the
config makes this possible.

But I think I will talk with you of list about this issue if you want
;-)

>Currently, we have done only 10-15 K clients (added to multi-handle),
>that we believe due to our HW limitations. Total number of clients,
>that is used may be several times higher due to the clients doing their
>after url fetching "sleeping" timeouts.

?! on one Computer(Server/PC)?!
Sorry I don't understand this sentence.

BR

Aleks
Received on 2007-05-04