Depends on the hardware - GSR's have different MTU's than 7600's for
example (and dependant on linecard too). We use 9216 between 7206VXR
and 7606 for example.
No, the change is immediate - "show interface" will tell you among other
commands...
KanREN runs Foundry (Brocade) NetIron XMR 4000's as our primary core infrastructure with an MTU of 9216. To make the change (this is Foundry-specific), we have to change some system-max settings which only take effect once the device has been rebooted (or at least it DID require a reboot in the IronWare 3.3.x days). It does NOT reboot immediately so you're free to make the change then perform a reboot at a convenient time.
That depends on the hardware. I've seen gear running as low as ~8k. I'd have to consult standard, but I think the max is 10k (10240).
Keep in mind the switch is not the only device on the network with jumbo frame limits. The NICs in your servers will also have limits.
Do I need to reboot the switch to take effect after
setting up it?
Again, this depends on the system. Many accept the change immediately, while others have to rebooted or interfaces reset to effect the change.
if it doesn't need to reboot, How can I know the
switch is running fine in this mtu 9000? eg: cisco
any tools to check?
"ping" will do. Set the packet size larger than the normal MTU (1500) and see if it crosses the network intact. If it's not working, A) the packets will be dropped, and B) the "oversized frame" counter (among others) should be clocking errors.
Most modern systems do that already (part of path MTU discovery.) And if there are no routers in the path (only the switch in question), then there's nothing to fragment it anyway.
None that have ever been accepted. From a quick google for manufacturer support, 9216 looks like the most popular number. But, as I said, it boils down to the largest frame *every* device on the LAN will accept. If there is a single device that only supports "9000", then that's your limit. And if there's a single non-JF device in the LAN, it throws a wrench into the whole thing. (This appears to be one of the sticking points as to why IEEE won't accept the addition of JF to any specs.)
--Ricky
PS: The topic pops up again with super-jumbo frames in 10G networks.
You might want to consider what the Internet/2 folks are doing because they
have a few years of experience with Jumbo frames. This page < http://noc.net.internet2.edu/i2network/jumbo-frames.html> gives an overview
and links to several presentations worth reading.
They are particularly useful for customers who want to send data between
data centres because they usually have the ability to set a large MTU on all
interfaces on their side.
Which standard are you referring to? AFAIK, nothing above 1500 is
standardised
I've had two different kinds of customer requests for jumbo frames
- customers that want very large frames for performance reasons;
Many ethernet switches support 9000 or more, some don't,
and some technologies like ATM support ~4470.
Sometimes the ability to provide them depends on tunnel modes.
- customers that want frames that are at least ~1700-1800 bytes
so that a few layers of IPSEC or VLAN headers or whatever
won't break the 1500-byte packets inside them.
* sam_mailinglists@spacething.org (Sam Stickland) [Tue 03 Feb 2009, 13:04 CET]:
For what it's worth, TCP will negiogate MSS and will work with mismatched MTU in a single LAN segment.
No
Machine 1 -- switch with 1500 byte MTU -- switch with smaller MTU -- switch with 1500 byte MTU -- machine 2
Same situation as when you have IP routers with smaller MTUs in the path that also do not send ICMP Fragmentation Needed errors (or those are dropped on the way to you)
If you configure one of those machines with an MTU equal to or smaller than the smallest MTU in the path then yes TCP (assuming MSS option is used) won't send packets that happen to be too big, but again, same story as for routers vs on a LAN. The problem isn't that machine 1 and 2 in the above example disagree on MTU, the problem is that equipment in the path disagrees on the MTU and cannot (in the case of switches) send notifications of such, or those will not arrive (in the case of stupid firewall admins in control of networks).
* sam_mailinglists@spacething.org (Sam Stickland) [Tue 03 Feb 2009, 13:04 CET]:
For what it's worth, TCP will negiogate MSS and will work with mismatched MTU in a single LAN segment.
No
Machine 1 -- switch with 1500 byte MTU -- switch with smaller MTU -- switch with 1500 byte MTU -- machine 2
Same situation as when you have IP routers with smaller MTUs in the path that also do not send ICMP Fragmentation Needed errors (or those are dropped on the way to you)
If you configure one of those machines with an MTU equal to or smaller than the smallest MTU in the path then yes TCP (assuming MSS option is used) won't send packets that happen to be too big, but again, same story as for routers vs on a LAN. The problem isn't that machine 1 and 2 in the above example disagree on MTU, the problem is that equipment in the path disagrees on the MTU and cannot (in the case of switches) send notifications of such, or those will not arrive (in the case of stupid firewall admins in control of networks).
Sorry, I should had clarified. I meant mismatched host MTUs within a jumbo MTU supporting L2 domain.