Do Active and Standby L4 Devices Both Send RST on TCP Health Check Failure?

Forums Technical Forums
,
1

Hello.

I have a question regarding the operation of F5 devices, and I would like to ask for your guidance.

1. If a TCP health check fails, do both the Active and Standby L4 devices each send a Reset (RST) packet?

2. Is it the normal behavior for both Active and Standby L4 devices to send Reset packets individually?

I would appreciate it if you could provide any relevant technical documentation for reference.

2

Hi Amit_Kumar9602,

About your questions

  1. Both devices in the HA group have individual monitoring using the non-floating IP, so yes, it is normal for each device to send the reset after the monitoring fails.
  2. It is normal behavior, both devices are monitoring the servers with their non-floating IP, take in mind that if a SW fails, the other device could take the active status to continue with the service.

I hope this clarifies your questions.

3

Hello

Thank you very much for the quick response. I have additional questions:

  1. Is there a technical document explaining the rationale behind the answer?
  2. Which part of the reference document I provided should I check to find the answer related to non-floating IP?
  3. In an Active-Standby setup, does sending an RST (Reset) impact the service session? I would like to know if it only resets the health check session or if there are other effects.

4

Whether both active and standby Layer 4 (L4) devices send TCP RST packets in response to a health check failure depends on the configuration of the devices and the failover or load-balancing mechanism they use.

5

Could you explain in more detail? Or is there any reference document I can check? Is it possible that no RST packet is sent even if non-floating IP is used?

6

Hi Amit_Kumar9602,

for your new questions:

  1. 404 Page not found | BIG-IP Documentation
  2. myF5
  3. The active device is the only one that processes the traffic for your applications if the service is down and the monitor marks the service down, the status of the VS will become offline and it restarts all connections to avoid send users to unresponsive servers.