Nginx-0.6.21

Changes with nginx 0.6.21 03 Dec
2007

*) Change: if variable values used in a "proxy_pass" directive 

contain
IP-addresses only, then a “resolver” directive is not mandatory.
resolver

*) Bugfix: a segmentation fault might occur in worker process if a
   "proxy_pass" directive with URI-part was used; bug appeared in
   0.6.19.

*) Bugfix: if resolver was used on platform that does not support
   kqueue, then nginx issued an alert "name is out of response".
   Thanks to Andrei Nigmatulin.

*) Bugfix: if the $server_protocol was used in FastCGI parameters 

and a
request line length was near to the “client_header_buffer_size”
directive value, then nginx issued an alert “fastcgi: the request
record is too big”.

*) Bugfix: if a plain text HTTP/0.9 version request was made to 

HTTPS
server, then nginx returned usual response.

Hi, all!

I have using stable nginx in one cluster with 6 nodes, using ip_hash.
I have some POSTs very slowly.
I have seen that some times the same POST request jump until 6 times
repeating the same operation, one by every cluster node. Some timeout,
I suppose, and jumping to next server.

How could I avoid that? Anyway to disable that timeout in that conf?

Regards,
Agustin

Hello!

On Mon, 3 Dec 2007, Agustin L. wrote:

I have using stable nginx in one cluster with 6 nodes, using ip_hash.
I have some POSTs very slowly.
I have seen that some times the same POST request jump until 6 times
repeating the same operation, one by every cluster node. Some timeout,
I suppose, and jumping to next server.

How could I avoid that? Anyway to disable that timeout in that conf?

http://wiki.codemongers.com/NginxHttpProxyModule

See proxy_next_upstream, proxy_send_timeout, proxy_read_timeout
directives.

Maxim D.

Thanks!

That resolve my problem.

Regards,
Agustin

Maxim D.
escribió:> Hello!