Red Hat Training

A Red Hat training course is available for Red Hat Satellite

6.4. Testing the Configuration

This section discusses basic testing of your load balancer and proxy configuration.
Initial Testing

Start the load balancer and the proxy machines. Note that the Satellite Proxy logs will not be created until the first requests are delivered (/var/log/rhn/). Try to install and remove an RPM file, such as zsh.

The /var/log/squid/access.log file on the load balancer should contain information similar to the following: 
1377540630.159
97 192.168.100.19 TCP_MISS/200 1515 POST https://lb.example.com/XMLRPC - ROUNDROBIN_PARENT/proxya.example.com text/base64
1377540630.733
529 192.168.100.19 TCP_MISS/200 1409 POST https://lb.example.com/XMLRPC - ROUNDROBIN_PARENT/proxyb.example.com text/xml
1377540639.968
87 192.168.100.19 TCP_MISS/200 3742 POST https://lb.example.com/XMLRPC - ROUNDROBIN_PARENT/proxya.example.com text/xml
1377540644.273
83 192.168.100.19 TCP_MEM_HIT/200 2238956 GET https://lb.example.com/XMLRPC/GET-REQ/rhel-x86_64-server-6/getPackage/zsh-4.3.10-5.el6.x86_64.rpm - NONE/- application/octet-stream
1377540646.765
100 192.168.100.19 TCP_MISS/200 1515 POST https://lb.example.com/XMLRPC - ROUNDROBIN_PARENT/proxyb.example.com text/base64
1377540647.291
485 192.168.100.19 TCP_MISS/200 1409 POST https://lb.example.com/XMLRPC - ROUNDROBIN_PARENT/proxya.example.com text/xml
This example shows that the client requests are balanced between the two different proxies. A single yum package install is still shared between the multiple proxies because multiple requests are involved during a package installation. The client IP address is 192.168.100.19 and you can see the named proxya.example.com and proxyb.example.com cache peers corresponding to the two different Satellite Proxies.
Squid Load Balancer Log Files

Refer to the following log files to monitor load balancer activity:

  • /var/log/squid/access.log for watching requests arriving from the client
  • /var/log/squid/cache.log for debugging SSL startup issues and cache peers on ports 80 and 443

You can use the netstat command to ensure that the squid load balancer is listening on the correct ports: 
# netstat -tulpn | grep squid | grep tcp
tcp        0      0 :::80               :::*           LISTEN      29120/(squid)
tcp        0      0 :::443              :::*           LISTEN      29120/(squid)