ARP Failure on LTM
I ain't a master but if you find this silly, please forgive me.
BIG IP <-> Internal VLAN <-> Cisco SW's <-> Servers
We have multiple Self IP's created for Internal VLAN so BIG IP can choose anyone to reach the Pool Members. We don't have SNAT enabled but NAT is enabled. ( it was already there when I checked the configuration)
Now, servers are using one of the self Ip as gateway IP. When ever we ping from outside network to those servers, I can see traffic hitting one of the self IP's but after that it times out. It works sometimes though..I don't understand why it is so intermittent. i did few captures but found that F5 was sending out ARP broadcasts but there wasn't any reply. Its more like (stacked switches in between) a direct connection from F5 to those servers.
all the Self IP addresses are part of same internal vlan and is connected to same 1.2 interface so no matter what self ip it uses, it will be going via 1.2 itself.
Do you think its because of arp time out? Default is 2seconds and then new request is broadcasted. Even if it, atleast for 3rd or 4th packet, it should get a reply isn't it?
I had a similar problem previously and I just used SNAT Automap and it worked. Stange..!!! But I can't test it out unless I am sure that it would work. (big people around and can't afford any outages etc).
please provide me your valuable suggestions.
Thank you.
Jack