opening sockets and cpu usage

[Pierre Sahores]

Hello,

Le 10 août 04, à 22:59, Dar Scott a écrit :

>
> On Aug 10, 2004, at 2:26 PM, Dar Scott wrote:
>
>> It looks like a bad thread is running when any socket is open.  It 
>> might be polling.
>>
>> This looks like bugzilla food to me.
>
> I don't see any new threads added.  A polling thread might be OK under 
> some conditions.
>
> Normally Rev makes 200 system calls a second, but with sockets on, it 
> makes about 80,000.

... and 1/20 of the sended requests, using the "open socket" command 
don't never arrive to destination... This is an Metacard historical 
bug. As a workaround, i avoid to use this command and use the "POST 
Method" to the Apache port + a php sockets listener/translator instead, 
where the php script is binded, first side, to Apache on port 80 and, 
second side, to the Revolution server-side app on the port xxxx this 
app accept connection :

> accept connections on port "xxxx" with message "newconnect"

Works just as expected and take advantage of the Apache+PHP ability to 
replace MC/Rev in what represants, in my opinion, one of its oldest and 
most dangerous MC/Rev limitation...

The php script example can be found in the list archives in searching 
about the Metacard+PostgreSQL threads.

Best, Pierre

> Dar Scott
>
> _______________________________________________
> use-revolution mailing list
> use-revolution at lists.runrev.com
> http://lists.runrev.com/mailman/listinfo/use-revolution
>
>
-- 
Bien cordialement, Pierre Sahores

100, rue de Paris
F - 77140 Nemours

psahores+ at +easynet.fr

GSM:   +33 6 03 95 77 70
Pro:      +33 1 64 45 05 33
Fax:      +33 1 64 45 05 33

WEB/EAI services & ACID DB over IP
"Mutualiser les deltas de productivité"