Re: About the web polygraph to test squid performance

tor 2007-05-31 klockan 20:19 +0800 skrev Leon Xiang Li: 
> Hi all:
> 
> 
> I set up three server :  polysrv  / polyclt / squid
> 
> I use polymix-1.pg to test the squid preformance.
> 
> But there is no much document available about the meaning of pgl
> configuration file and working theory of polygraph.

There is quite a bit of documentation available from
http://www.web-polygraph.org/docs/

There is not much documentation specific to polymix-1 as there only is
one single workload parameter in polymix-1 namely the request rate.

For a more realistic simulation the polymix-4 workload is recommended.
But it's also considerably more complex to configure..

> I run polysrv as:
> 
> ./polysrv --config /usr/local/polygraph/polymix-1.pg --verb_lvl 4

You haven't specified the port number..

> I run polyclt as:
> 
> ./polyclt --config /usr/local/polygraph/polymix-1.pg --verb_lvl 4
> --proxy 192.168.16.161:8080 --log /tmp/client.log
> 
> 
> I change the polymix-1.pg as following:
> 
> addr[] srv_ips = ['polyserverip:8080', 'polyserver:8888' ];
> addr[] rbt_ips = ['polyclientip' ];

Try just with the polyserverip:8080, there should only be one srv_ips
entry per polygraph server in polymix-1.

> But after 1 hour, it still runing and in i-warm phase.

Probably the client thinks there is two servers, when there only is
one..

Regards
Henrik

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)

iQCVAwUARl8Ta0NPQ5Kbx8daAQJCzQQAoWW2fX1ZiYWo1lava9q2vzVzy8Hz9HTM
NQjuibIvHu8NmoZoKx3wiy6FrNFjvMHUu98vpNa27bApa2z/I7VekK8h1qwjgsTS
Kl5XH8uCYQNLzoCH67bZSDsDkfAdwjxD/UmsEPLs6r/TlJS3tBSrlos8WKbOzpNL
j7n/PdGiqJ8=
=Bagg
-----END PGP SIGNATURE-----