https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=291100

Zhenlei Huang <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[email protected]

--- Comment #3 from Zhenlei Huang <[email protected]> ---
(In reply to Achilles Gaikwad from comment #0)

> ## issue:
> The current 20-minute ARP cache timeout, while appropriate for traditional 
> stable
> network environments (Like on-prem systems), causes connectivity issues in 
> modern
> cloud environments. It is not common that in on prem the IP pool gets recycled
> quickly within 20 minutes. Where as in cloud environments - 

> 1. Short-lived instances: Client systems may operate for less than 20 minutes 
> before
> termination, the IP address that the virtual machines held will be reused by 
> other
> virtual machines.
> 2. Due to [1] when IP addresses are quickly reassigned to new virtual 
> machines, the
> OS continues attempting to communicate with the cached (now stale) MAC address
> instead of performing a fresh ARP resolution until 20 minutes.

When the IP address is assigned to new a VM's interface, the VM will broadcast
ARP request to detect duplicated IP,

```
# jexec 2 tcpdump -envvvi epair1a
tcpdump: listening on epair1a, link-type EN10MB (Ethernet), snapshot length
262144 bytes
10:15:36.179644 58:9c:fc:10:7c:d3 > ff:ff:ff:ff:ff:ff, ethertype ARP (0x0806),
length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 172.16.1.99 tell
172.16.1.99, length 28
```

and the neighbors can update their ARP cache entry to reflect the new IP-to-MAC
mapping,

```
# tail -n1 /var/log/messages
Nov 20 10:15:36  kernel: arp: 172.16.1.99 moved from 58:9c:fc:10:ff:3c to
58:9c:fc:10:7c:d3 on epair0a
```

So it does not matter how long the ARP entry lifetime is.

Do you have troubles that the ARP cache entries are stale ?

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to