For Solaris 10, I've got the answer I needed at :
I would like to do the same network teaming/bonding (as it's
termed in Linux) for HP-UX & RHEL & older version of Solaris 9.
On HP9000, HP-UX version :
B.11.11 U 9000/800
On Itanium, HP-UX version :
On Redhat, version is (heard network teaming in Linux is called bonding) :
version 9 (unlike ver 10, heard that this needs a trunking software)
Appreciate if the step by step commands, the configuration
file (& directory where it's located) are given. If any addtional
software is needed (say trunking or 802.x aggregation), pls
provide the location to download from.
Priority is to provide redundancy in case one link/NIC is down
but would be nice if load sharing of traffic is available for the
If possible, avoid IPMP (as this requires 3 IP addresses) &
firewall rules may be affected. A single IP address is best,
if not, then suggest the next best option for each of the OS
If you're familiar with only one or two of the above OS, your
answers are appreciated too.
If there's any specifics on the Cisco switch (6513) end that
needs special handling, kindly highlight as well
JustUnix answer is almost complete, just that it's missing the HP-UX part.
Do you happen to know about the HP-UX part & if possible Cisco
I've got an answer from a member of this forum that link aggregation
on Cisco can only be performed if both the links from the server are
connected to a common switch chassis. If it's two different switches,
this is not feasible on the switch's end.
The question is :
suppose I simply connect both links from the server to two separate
switches' chassises (no LACP/link aggregation setup on switch end,
just assign VLAN only), will this still provide 'seamless' redundancy
(load sharing is not crucial to us)? Seamless in a way that when one
switch sufferred power trip, the user will not see a disruption while
accessing the server.