Re: DoS issue
"Étienne KASPERCZYK" <etienne-bAYckG+VU/[email protected]>
| Newsgroups | gmane.comp.web.fastcgi.devel,gmane.spam.detected |
|---|---|
| Message-ID | <[email protected]> |
I don't really see what you are asking for ? If you want the request to be answered, just raise the number of max process, you can even disable the limit (never tested actually, but i don't see why you couldn't). By the way, for lighty, the "flag" within the load balancer already exists, since there's a load counter for each process. The mod_fastcgi knows exactly the load for each process. On 8/15/07, Arvid Picciani <[email protected]> wrote: > > On Wednesday, 15. August 2007 15:04:01 Étienne KASPERCZYK wrote: > > By "server side busy flag" you mean : an appropriate error code > displayed > > to the end user (which is : the web browser) ? > no. a flag within the load balancer indicating... load. > > I've never seen/read anything like this in lighty, but AFAIK there's a > > "load" integer for each process, and until it doesn't have reached the > > maximum, it'll just wait. So i guess you'll need to reach the > ((max_process > > * max_load) + 1)th connection to see an error appear. > nah i don't want any error. even if it would display an error that's even > more > a DoS. > the problem is that any aditional requests are ignored instead of starting > another instance that can answer those > > > -- > best regards / Mit freundlichen Grüßen > Arvid Picciani > -- > DO NOT SEND ME POWERPOINT(tm) OR WORD(tm) DOCUMENTS! > pdf or any free as in free speech format is fine. > ___________________________________ > fastcgi-developers mailing list > http://fastcgi.com/fastcgi-developers/ > ___________________________________ fastcgi-developers mailing list http://fastcgi.com/fastcgi-developers/