# health check service testing vs host logging **URL:** \ **Category:** aFleX\ **Tags:** health-monitor\ **Created:** [September 13, 2021, 12:19pm UTC](https://community.a10networks.com/t/health-check-service-testing-vs-host-logging/676 "2021-09-13T12:19:30Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![bseppanen](https://avatars.discourse-cdn.com/v4/letter/b/5e9695/32.png) [@bseppanen](https://community.a10networks.com/u/bseppanen)\ **Post date:** [September 13, 2021, 12:19pm UTC](https://community.a10networks.com/t/health-check-service-testing-vs-host-logging/676/1 "2021-09-13T12:19:30Z")
We are in the process of load balancing radius with ACOS 4.1.4-GR1-P2 with Direct Server return. The load balancing works well, and now I have some awkward consequences I need to deal with. The failover works when a health check is enabled and marks the service as offline if the service is not functional. so we need a service check sent to port 1812/1813 for radius. each health test of a packet sent to udp port 1812/1813 results in Windows NPS logging “Malformed Radius Packet Received”. Service testing should pick up a failure as soon as possible so even if the testing was delayed to once every 30 seconds for 2 services, we’re talking 7200 logged messages over an hour. 7200 inconsequential logged messages. What best practices, or coping strategies have others utilized in the past to offset issues such as this? One coping mechanism I want to implement is just make one health check dependent upon the other. Does anyone have pointers or references on how health check dependencies can be configured? TIA ---
**Author:** ![bseppanen](https://avatars.discourse-cdn.com/v4/letter/b/5e9695/32.png) [@bseppanen](https://community.a10networks.com/u/bseppanen)\ **Post date:** [September 13, 2021, 2:44pm UTC](https://community.a10networks.com/t/health-check-service-testing-vs-host-logging/676/2 "2021-09-13T14:44:08Z")
# [https://a10support.force.com/customers/kA160000000CeRI?srPos=1&srKp=ka1&lang=en\_US](https://a10support.force.com/customers/kA160000000CeRI?srPos=1&srKp=ka1&lang=en_US) unfortunately the consequence of the available options still results in port 1813 being tested twice if I make that the port we test and use override-port 1813 for port 1812. Was hoping that it would be possible to re-use the results of test 1813 to update 1812, without it testing 1813 for a second time. ---
**Author:** ![ebellamy\_a10networks.com](https://avatars.discourse-cdn.com/v4/letter/e/a698b9/32.png) [@ebellamy\_a10networks.com](https://community.a10networks.com/u/ebellamy_a10networks.com)\ **Post date:** [September 13, 2021, 2:54pm UTC](https://community.a10networks.com/t/health-check-service-testing-vs-host-logging/676/3 "2021-09-13T14:54:25Z")
Hi, It Eddie Bellamy from A10 UK SE team. please email me @ [eddie@a10networks.com](mailto:eddie@a10networks.com) will be glad to help you will a solution. I will post the answer on here when done. best regards, Eddie ---
**Author:** ![ebellamy\_a10networks.com](https://avatars.discourse-cdn.com/v4/letter/e/a698b9/32.png) [@ebellamy\_a10networks.com](https://community.a10networks.com/u/ebellamy_a10networks.com)\ **Post date:** [September 13, 2021, 3:12pm UTC](https://community.a10networks.com/t/health-check-service-testing-vs-host-logging/676/4 "2021-09-13T15:12:58Z")
Hi. incase you are busy, you can combine health monitors and timers for each using a Compound Monitor. see command syntax below. (this was taken from the ACOS 5.20 SLB.pdf user guide. Configuring Compound Health Monitors – Timeout (CLI Example) For a compound health check configured as follows: ACOS(config)# health monitor monitor1 ACOS(config-health:monitor)# interval 3 timeout 3 ACOS(config-health:monitor)# retry 2 ACOS(config-health:monitor)# method tcp port 80 ACOS(config-health:monitor)# exit ACOS(config)# health monitor monitor2 ACOS(config-health:monitor)# interval 5 timeout 5 ACOS(config-health:monitor)# retry 1 ACOS(config-health:monitor)# method tcp port 8080 ACOS(config-health:monitor)# exit ACOS(config)# health monitor compound-monitor ACOS(config-health:monitor)# interval 6 timeout 6 ACOS(config-health:monitor)# retry 3 ACOS(config-health-monitor)# method compound sub monitor1 sub monitor2 and GUI details below. ![image.png](https://us1.discourse-cdn.com/flex002/uploads/a10_community/original/1X/41cc071d03253970b8524b2c8621c7d504d5d510.png)