We need your help to fix our radius server error. I kindly request you to help us on the same.

We are facing ‘Child is hung’ message in our radius server and we have to restart radius server to become normal.

As per our analysis the error message comes whenever there are more load on the server.

Configuration:

max_request_time = 30

cleanup_delay = 5

max_requests = 12800

max_queue_size = 131072

start_servers = 5

max_servers = 32

min_spare_servers = 3

max_spare_servers = 10

max_requests_per_server = 0

Our system TPS is 100 (at max 120). Since we have kept max_request as 12,800 and max_request_time as 30 seconds, we will be able to support max backlog 3000 (TPS x max_request_time).

That is our system can clear first 3000 requests in the queue before it faces ‘time expire’ (max_request_time). However we have set max_request as 12,800 during high load, system is facing ‘child is hung’ when the backlog grows more than 3000.

Is our above understanding correct? If it is correct, we can resolve our issue by setting max_request as 3000 i.e. during high load, we will not face ‘child is hung ‘ and no need to restart radius server to become normal.

Looking forward for your reply.