Forum: NGINX SPDY giving HTTP 500

I've compiled nginx with the SPDY patch, at the current latest, so nginx
1.3.14. As far as I can tell, the make was fine, no errors. I'm 
compiling
against openSSL of the system: OpenSSL 1.0.1c 10 May 2012, on my CentOS 
6.3
server (OVH Kernel). Server has ok amount of RAM, at least 20GB is free.

./configure was with minimal options.

However, when a client attempts to connect over SPDY, they are faced 
with a
HTTP 500 error. The errors logs aren't the most useful,. indicating 
memory
issues, and I can't even find the error when the logging is set to 
debug.

Here's the error log snippet pertinent:

2013/03/11 22:32:08 [emerg] 10589#0: *1 malloc(18446744073709551615) 
failed
(12: Cannot allocate memory), client: ****, server: ****, request: "GET 
/
HTTP/1.1", host: ****

Any ideas why nginx is trying to allocate so much memory?

Posted at Nginx Forum: 
http://forum.nginx.org/read.php?2,237244,237244#msg-237244

On Tuesday 12 March 2013 01:44:01 ct2k7 wrote:
>
> Here's the error log snippet pertinent:
>
> 2013/03/11 22:32:08 [emerg] 10589#0: *1 malloc(18446744073709551615) failed
> (12: Cannot allocate memory), client: ****, server: ****, request: "GET /
> HTTP/1.1", host: ****
>
> Any ideas why nginx is trying to allocate so much memory?
>

Could you show the debug log?

  wbr, Valentin V. Bartenev

--
http://nginx.org/en/donation.html

On Tuesday 12 March 2013 01:44:01 ct2k7 wrote:
>
> Here's the error log snippet pertinent:
>
> 2013/03/11 22:32:08 [emerg] 10589#0: *1 malloc(18446744073709551615) failed
> (12: Cannot allocate memory), client: ****, server: ****, request: "GET /
> HTTP/1.1", host: ****
>
> Any ideas why nginx is trying to allocate so much memory?

This bug was fixed in:
http://nginx.org/patches/spdy/patch.spdy-68_1.3.14.txt

Thank you for the report.

  wbr, Valentin V. Bartenev

--
http://nginx.org/en/donation.html