Not working with nginx 1.0.2

[GoogleCodeExporter]

Compiled with nginx 1.0.2.

First if you dont have default configfile on filesystem, you are getting this 
error message from every http request:

[error] 3711#0: *1 stat(/etc/hsflowd.auto) failed while logging request

If you make 0 byte /etc/hsflowd.auto, nginx will segault on every http request

Then if trying with example configuration i am getting error:
[error] 3935#0: *2755 line=rev_start tokc=^@rev_start while logging request

So it seems sflow module (0.9.3) does not work atm. 

Original issue reported on code.google.com by ppan...@gmail.com on 18 May 2011 at 12:24

[GoogleCodeExporter]

Thanks, I'll check it out.  What OS are you running on?

Installing hsflowd might result in the format it is looking for in 
/etc/hsflowd.auto:
http://host-sflow.sourceforge.net

Neil

Original comment by neil.mck...@gmail.com on 27 May 2011 at 2:14

[GoogleCodeExporter]

CentOS 5.6 64-bit. 

Original comment by ppan...@gmail.com on 28 May 2011 at 5:56

[GoogleCodeExporter]

The first problem is that the example in the README has the line "agent eth0" 
when it should have "agentIP <ipaddress>".  Running hsflowd would take care of 
that,  but I should fix the README.

The second problem is that the error logging to complain about that uses 
nginx_log_error() and I think I may have the wrong arg specifiers.  That's 
where the segfault is coming from.

The third problem is an extra line that was only there for debugging and should 
have been removed.  That's the reason for the error log messages that you are 
actually seeing.

I'll clean this up and post a new version.  Thanks for your help.

Neil

Original comment by neil.mck...@gmail.com on 29 May 2011 at 10:59

• Changed state: Accepted

[GoogleCodeExporter]

Posted version 0.9.4 which should fix this.

Neil

Original comment by neil.mck...@gmail.com on 30 May 2011 at 10:24

[GoogleCodeExporter]

Fix verified. Closing ticket.

Original comment by neil.mck...@gmail.com on 27 Jun 2011 at 10:56

• Changed state: Fixed