Forum Discussion
NimbostratusQuestion about VIP
Hello,
So we ran into this issue where we had a forwarding VIP which was source to one destination. One day we had to reboot same server. Once the server was rebooted and when it came back up but it was complaining about IT conflict.
So after hours sending on this finally figured out it was the VIP causing this issue so had to disable to corresponding VIP, reboot the server again and this time it did come backup with correct IP and without any complaining. Once server is up I went in and enabled the VIP.
But is this normal behavior for F5 is there any work around?
I know there is a an alternative for this solution i.e use standard VIP with pool member.
Thanks
5 Replies
- Kevin_Stewart
Employee
I think the biggest question would be how you have all of this set up? What does your forwarding VIP look like? Are you using a duplicate IP somewhere? And to your last question, is there a specific need for a forwarding VIP versus a standard VIP?
- Kevin_Stewart
It almost sounds like an IP conflict. Is the server also using the 192.168.100.20 address? Where does the server sit in relation to the VIP? Understand that the Source value in the VIP is the network from which the VIP which accept traffic, and the destination is the IP of the VIP itself. For a standard VIP you still need a pool to send the traffic to.
Yaoxie_117356is there need to enable the vip on the vlan facing server side? Normally enable it on the vlan facing client should be enough.
mr_evil_116524Hmm... I think I might know the issue here, So far I have only noticed this happening on a Physical device connected to F5. And this physical server have 2 nic and they are both are teamed. So this is what I suspect happened when the server when down in F5 ARP table there was an entry for that server with a MAC access, how ever when the server came back up it got a different MAC access hence that didnt match the ARP table in F5. However we have rebooted VM many many times and I do have VIP for source to one destination and didnt see this happening. Can anyone confirm what I am thinking can happen?
- mr_evil_116524
We have found the resolution for this issue, you will need to unchecked few boxes by going to Local Traffic --> Virtual Servers : Virtual Server List --> Virtual Address List (2 tab from the top) and select the IP in question and turn off ARP and ICMP Echo.
Once you do that and restart you back end node and go to Network --> ARP --> Dynamic list you will see the IP in question will have status incomplete and when it box fully up that ARP table will resolve itself.
Hope this can be useful to some people out there. And Thanks to F5 for showing us this config.
