A Unicenter AutoSys r11 remote agent responds most of the time but periodically times out. Looking for recommendations.
General network issues should be investigated by the network admin. They can run diagnostics such as ping, nslookup, traceroute or tracert and examine the paths and response times between the hosts and network performance.
From an AutoSys perspective, some timeouts may be caused as a result of too much work being asked of an agent at a given time and the resources have been exhausted. An example, trying to start 5000 jobs on 1 remote agent, at 1 time. The machine might not have the available cpu or memory to allow all these requests to come in simultaneously, or in a timely manner. It is recommended to stagger the starting of jobs some, so as not to flood a system. That will most often resolve those type of timeout issues.
Additionally, one may try setting the environment variable "AGENT_RECV_TIMEOUT" on the scheduler host prior to starting the scheduler. This allows for the configuration of the timeframe before a request to an agent would be considered timed out. By default this environment variable is not set. Without the variable set, the scheduler will use a default value of 21 seconds.