You might experience difficulty when you configure High Availability in your environment. You can use a variety of procedures for diagnosing and fixing these problems.
- Log in to Unified Access Gateway console.
- Run ip addr command to check if the configured virtual IP address is assigned to eth0 interface.
- Ensure virtual IP address is assigned within the same subnet as eth0 interface. Ensure it is reachable from the client machine. If there are connectivity issues then it could be due to virtual IP address not being unique and already assigned to a physical or virtual machine.
- In the haproxy.conf file in log bundle, configuration related to the current cluster is available. For example,
server uag1 127.0.0.1:XXXX ..... server uag2 <IP of machine 2>:XXXX .... server uag3 <IP of machine 3>:XXXX ....The back-end configuration is based on the settings configured on Unified Access Gateway
- lb_esmanageris for Horizon and Web reverse proxy use cases.
- lb_cg_server is for Content Gateway use cases.
- lb_tunnel_server is for Tunnel use cases.
- In the haproxy.conf file in log bundle, you can find details about the client connection source, corresponding connection sent, and the Unified Access Gateway server that handles the connections. For example,
2018-11-27T07:21:09+00:00 ipv6-localhost haproxy: incoming:::ffff:<IP of Client:xxxx> backend:lb_esmanager connecting-server:uag2/<IP of uag2> connecting-through:<IP of primary node:xxxx> wait-time:1 connect-time:0 total-incoming:1 total-outgoing:1 total-to-server:1
To view the statistics, see Collecting Logs from the Unified Access Gateway Appliance.
Table 1. Example of a CSV File Column Name Description scur Indicates the current number of concurrent connections handled by this server. smax High watermark of concurrent connections handled by this server during current uptime. stot Indicates the total number of connections handled by this server during current uptime. bin Indicates the total number of bytes sent to this server. bout Indicates the total number of bytes received from this server. status Indicates the status of the server. For example, if it is up or down. This is based on the last health check performed on this server.
Multiple primary node election issues can be seen in the following cases,
- Different group ID or virtual IP address configured on the nodes that are intended to form the cluster.
- Virtual IP address and eth0 in different subnet.
- Multiple NICS on Unified Access Gateway configured within the same subnet.