MPLS MTU

We are facing serious problems related to MPLS MTU. Due to fragmentation some of the applications are not working. I have enabled the mpls mtu 1508 across the path of MPLS but some of our switches do not support jumbo frames.
The strange this is that we are not only facing this issue with customer which are connected with us via Ethernet (wireless and cable) but also customers having WAN connection like ATM.

Hi Ahmede...
We had a similar issue with Windows 2000/2003 platforms. In our case, it was a Microsoft bug which a patch fix it... However, we don't manage the Microsoft servers, so we decided to pursue this with a solution in the router. Fragmentation was the main cause, and fixed the problem by adding the following command to the interfaces :
ip tcp adjust-mss 1272
We are still working to identify why '1272', but if you ping from some PCs you can have a payload of up to 1400+, but from other workstations (mine is an XP) I could do only 1272. Hope this helps.

Similar Messages

Layer 3 Etherchannel and MPLS MTU

Greetings,
If two 7609 are connected through layer 3 giga etherchannel and 7609-1 sends a frame with 1514 bytes plus two MPLS header to 7609-2, does the command mpls mtu 1522 under the giga etherchannel interface on both routers is enough?
Or it is necessary to add the global command system jumbomtu 1522 and the interface command mtu 1522 under physical gigainterfaces participating on the channel?
Cisco docs say many platforms will accept frames higher than 1500 bytes (up to 24 bytes) to accommodate control flags; that is why I am considering just mpls mtu xxx is enough.
Also, reading Cisco doc I understood that system jumbomtu xxxx and mtu xxxx should be used together. I tested sending frames higher than 1514 bytes (without MPLS tags) and system jumbomtu xxxx alone was not enough, ping failed; it was necessary to add mtu xxxx under the interface. Do you have a different experience?
Thanks,
Alaerte Gladston Vidali
IBM Global Services - SO
Tel.55+11+2121-2879 Fax:55+11+2121-2449

Hi
We had some peculiar exp with 7600 (7609/7606) boxes in which we did change the physical mtu of the interface instead of mpls mtu.
Again that did the trick for us even in 7200 where we monitored the same kinda behaviour..
regds

Confused about mpls mtu command

hi,
i confuse about mpls mtu command
test platforms are 76 pfc3b,mpls,gigabit sip400 spa interface
if i didn't config mpls mtu command ,using default,ping command is successful,if more than 1496 packets, i can see fragment from show ip traffic.
if i config mpls mtu override 1504,ping command is sucessful too. there is fragment too when i use 1501 byte packet.
if i config mpls mtu override 1524 byte.
ping command failed if i use packet more than 1500, , all packet are droped,even 1501 byte.only 1500 byte packet can success.
all config above interface mtu is 1500.
this confused me.
why i use default 1500 interace mtu, mpls mtu override 1504 ,ping packet can fragment,ping success, but i use mpls mtu override 1524, i can see fragment in show ip traffic,but ping command failed. i can't see packet in destination router,how this work.
thank you!
jun

topology is simple
7609-1--sip GE spa----7609-2--pos---7609-3--flexwan E1-----7604-1--ge--ce
i config mpls mtu 1524 between 7609-1 and 7609-2 . and keep interface mtu 1500 default.
ping packet from 7609-1 to 7604-1 loopback 0.
ping 1500 byte packet is ok, but ping 1501 byte packet is totally lost.then i config mtu 1524 between 7604-1 and 7609-3, it is useless,notwork, i can't see packet coming from 7609-1 on 7604-1.
but i add config mtu 1524 between 7609-1 and 7609-2. config mtu 1500 between 7604-1 and 7609-3,ping 1501 bytes from 7609-1 to 7604-1 loopback0 is ok. but i can see fragment from show ip traffic command in 7609-3.
i have a question, why we need mpls mtu command. if we don't change interface mtu,just only config mpls mtu 1524, it doesn't work, if we just change mpls mtu,how it work in the ios. if we config interface mtu 1524,interface mtu size is big enough, it seems mpls mtu command is useless, we don't need mpls mtu command, just interface mtu 1524 is ok.
why we need mpls mtu command. we just only change interface mtu is enough.
thank you!
jun

Pmtu and interface mtu,mpls mtu

Hi,
If pmtu work , do i need config interface mtu and mpls mtu on interface facing core,or p-p along the path.
thank you!

Not necessarily. However, Path MTU is not guaranteed to work all the time because of things like firewalls and access-list. From a performance perspective, you will want the maximum possible MTU, which is what makes it adviseable to change the MTU (IP/MPLS).

MPLS MTU on FE interfaces with fixed interface MTU

What is the recommendation for setting MPLS MTU when the hardware does not permit the setting of the interface MTU:
PE-Edge-2(config)#int fastEthernet 0/0
PE-Edge-2(config-if)#mtu 1524
% Interface FastEthernet0/0 does not support user settable mtu.
The MPLS MTU could be changed with "mpls mtu 1524", but this gives the error:
PE-Edge-2(config-if)#mpls mtu 1524
*Dec 24 20:54:43.467: %TFIB-3-MPLS_MTU_SET: Setting mpls mtu to 1524 on FastEthernet0/0 which is higher than the interface mtu 1500 . This could lead to packet forwarding problems including packet drops.
Is it permissible to set the MPLS MTU greater than the interface MTU, as I thought that MPLS packets cannot not be fragmented, but rather simply dropped?
In this scenario, is the only solution to limit the size of inbound packets with something like "ip tcp adjust-mss 1476" to avoid any potential packet drops?

if your doing AToM then you have to change the mpls mtu in order to transport full frames over your backbone. I have the same issue with the mtu and setting the mpls mtu to 9196 while leaving the mtu at 1500 did not impact regular traffic at all but it did allow our l2vpn's to function correctly.

MPLS MTU issue (diagram attached)

Dear Experts ,
I'm trying to run MPLS between two 7206 vxr router which is connected behind a Foundry Big Iron L3 router which is configured as a L2 device .
I put on my 7206 as
int gi0/1
ip address 172.16.14.1 255.255.255.0
mpls ip
mpls mtu 1526
But it seems that there is tag assiging ,but traffic like https didn't work after that . is there is any clue ?
Regards
Haris

Hi ,
There is no way to increase the MTU size on foundry . But it's giving me some error
BigIron1(config-if-e1000-1/8)#ip mtu 1526
Error - 1526 not between 576 and 1500
BigIron1(config-if-e1000-1/8)#mtu ?
Unrecognized command
How can know that whether the foundry supporting giant packets / not ? . I think it's not supporting giants , the following is the output of ping sweep from one cisco router to other one .
If it's not supporting giants what will happens to my mpls packets with size 1508 ? . Is it possible me to put the command "mpls mtu 1500" ,so that mpls mtu will limit to 1500
router1#ping
Protocol [ip]:
Target IP address: 213.12.90.1
Repeat count [5]: 1
Extended commands [n]: y
Source address or interface: 213.12.83.137
Set DF bit in IP header? [no]: y
Sweep range of sizes [n]: y
Sweep min size [36]: 1450
Sweep max size [18024]: 1550
Sweep interval [1]:
Type escape sequence to abort.
Sending 101, [1450..1550]-byte ICMP Echos to 213.12.90.1, timeout is 2 seconds:
Packet sent with a source address of 213.12.83.137
Packet sent with the DF bit set
Success rate is 45 percent (46/101), round-trip min/avg/max = 20/44/120 ms
the ping is same with and without DF bit set after enabling MPLS . But before enabling tag switching as I remember I was able to ping without DF Bit set
Regards
Haris

MPLS MTU size

Hello,
Please correct me if I am wrong. The MPLS MTU size is 1518 to 1520 and the normal MTU size is 1500, So does it means that while configuring the MPLS network on the Switches attached with the MPLS CE router we MUST have to configure the MTU size to 1520 or not.
Please confirm

MPLS is not configured between the CE and the PE. So if your IP mtu is 1500 then the switch between the CE and PE needs to support 1500 only.
The mpls mtu needs to be configured in the core where ethernet switches are used.
Hope this helps,

Applications fail when "mpls mtu 1526" is configured

Hi,
I'm facing an issue with an MPLS VPN backbone.
A customer's IP host at one location can ping and traceroute to application servers at it's central site at another location. The IP path traverses four P/PE routers, as follows:
IP Host (spoke site) -- CE1 -- P/PE1 -- P/PE2 -- P/PE3 -- P/PE4 -- CE2 -- Servers (central site)
However, access to all applications fails.
Network configurations as follows:
(1) All the P/PE routers are Cisco 7609, SUP720-3BXL, IOS 12.2(18)SXF6.
(2) All the core links are configured with "mpls mtu 1526".
(3) All the core links are GE, connected to SPA-2X1GE on a 7600-SIP-400 linecard. Some connect to the SFP port on the SUP720.
(4) Only some of the P/PE routers are configured with "ip tcp path-mtu-discovery".
After removing the command "mpls mtu 1526" from all core links, access to all applications and services works again.
Please advise if I've configured anything wrongly.
Thank you.
B.Rgds,
Lim TS

There are two ways to look at this,
1) First To solve the problem, on all your Core and Core Facings links configure "mtu 1526 or larger" (PE-P, P-P) mtu 1526 or larger.
2) Second "MPLS MTU 1526" command only helps in labelling the fragmented payload which is beyond 1526. Till then if anything is fragmented it doesnt label the fragments.
So now why does it work when its removed,
a) verify what is your actual link mtu.
b) "mpls mtu" command doesnt change the physical mtu it just tells the IOS when to label fragments. and the 1526 is the threshold to fragment the labelled packet and label the fragments also.
c) One important thing to note is even when you dont configure the "mpls mtu" command it is still there by default but now it uses the MTU value of the physical link, so if your link is of 1500 MTU the default "mpls mtu" is also 1500.!
d) also go through this previous thread for more information the same topic has been discussed at length.
MPLS: MTU Ethernet MPLS Network Sep 26, 2006
HTH-Cheers,
Swaroop.

Why we need mpls mtu command?

if interface mtu is not big enough to take mpls packet. we just increase interface mtu. why we need mpls mtu command.
if we just only increase mpls mtu.there is problem if mpls mtu biger than interface mtu. so it seem mpls mtu command is useless.
why we need mpls mtu command!
thank you!
Jun

Hi,
i test it in netowork,like this\
topology is simple
7609-1pe--sip GE spa----7609-2p--pos---7609-3p--flexwan E1-----7604-1pe--ge--ce
i config mpls mtu 1524 between 7609-1 and 7609-2 . and keep interface mtu 1500 default.
ping packet from 7609-1 to 7604-1 loopback 0.
ping 1500 byte packet is ok, but ping 1501 byte packet is totally lost.then i config mtu 1524 between 7604-1pe and 7609-3, it is useless,not work, i can't see packet coming from 7609-1 on 7604-1.
then i add config mtu 1524 between 7609-1pe and 7609-2. config mtu 1500 between 7604-1 and 7609-3,ping 1501 bytes from 7609-1 to 7604-1 loopback0 is ok. but i can see fragment from show ip traffic command in 7609-3.
i have a question, why we need mpls mtu command. if we don't change interface mtu,just only config mpls mtu 1524, it doesn't work, if we just change mpls mtu,how it work in the ios. if we config interface mtu 1524,interface mtu size is big enough, it seems mpls mtu command is useless, we don't need mpls mtu command, just increase interface mtu 1524 is ok.
why we need mpls mtu command. we just only change interface mtu is enough.
when i config mpls mtu override 1524,this is a warning in ios:
Setting the mpls mtu to 1524 on interface serial1/0/0:0, which is higher than the interface MTU
1500. This could lead to packet forwarding problems including packet drops.
You must set the MPLS MTU values equal to or lower than the interface MTU values.
thank you!
jun

MPLS MTU command

I need to change the MTU on a router 7507 interface (VIP4-80 card with PA-2FE) using the "mpls mtu 1508", but it is on a production network, so I need to know:
Does this command cause the router to reset the FE ports? or the VIP card? or everything is transparent to the router?
Please let me know because I have to do it ASAP.
Regards
Alex

"mpls mtu" or "tag-switching mtu" will not reset the VIP or the Interface (based on personal experience on 75xx series running 12.0 (23) Service Provider IOS) . You don't have to be concerned with resets. Obviously it has the potential to impact your MPLS traffic but increasing it can only help your cause. It certainly won't make thing worse. Good luck.

MPLS MTU on various ethernet controllers

Is there some documentation available which router ethernet interface controller supports which maximum frame size ?
Background:
We have some intra-PoP crossover FE connections in an MPLS VPN based network. To avoid fragmentation we'd like to increase the mpls mtu to e.g. 1520bytes. I know that this isn't an issue on a PA-2FE-TX (7206-VXR), but there are also c3745 used as PE / P boxes. Does anybody know what frame sizes the onboard FEs of the 3745 and the NMs of these platforms (e.g. NM-2FE2W) can support ?
There is a document available for switches on CCO:
http://www.cisco.com/warp/public/473/148.html
Something in that style whould be great ...
thanks,
Stefan

Any update on this ?

MPLS MTU Issue with PA-FE-TX

Dear All,
I got an ethernet backbone that do not support jumbo frame. The backbone connected my Cisco7206/NPE-G2 and Cisco7507 on each side.
The topology as below:
Cisco7206-Gi0/1<---Ethernet--->Cisco7507-F5/1/0.50 (I am using 802.1q sub-interface)
The config as below:
interface GigabitEthernet0/1
bandwidth 4000
ip address 10.86.40.2 255.255.255.252
no ip redirects
no ip proxy-arp
ip ospf network point-to-point
ip ospf cost 10
ip ospf mtu-ignore
load-interval 30
duplex full
speed 100
media-type rj45
negotiation auto
mpls label protocol ldp
mpls ip
mpls mtu 1508
end
interface FastEthernet5/1/0.50
encapsulation dot1Q 50
ip address 10.86.40.1 255.255.255.252
no ip redirects
no ip proxy-arp
ip ospf network point-to-point
ip ospf cost 10
ip ospf mtu-ignore
mpls label protocol ldp
tag-switching mtu 1508
tag-switching ip
end
interface FastEthernet5/1/0
description Trunk
no ip address
load-interval 30
full-duplex
end
The problem is when i using "ping vrf NM_TEST x.x.x.x size xxxx", the maximum size can only up to 1498byte. I have increase all of my switchs MTU to 1530, but also cannot ping with 1500byte. Is there any way can help to solve the problem?
Thanks for your kindly help!
Regards

Hi Mohamed,
Its all about the implementation, with AToM, VPLS, MPLS TE, MVPN and any relatively new technology that is still being developed and enhanced you must take care about the vendors' implementations and which standards and RFCs is it complying with and to what level. I've seen codes that do fragmentation for AToM and codes that don't, codes that support out of order recovery and codes that don't. But in the end the RFCs and standards are there, whether you are using a code that supports what you need or not is your call.
Well, taking the AToM fragmentation and reassembly for example, I quote from "RFC 4623 Pseudowire Emulation Edge-to-Edge (PWE3) Fragmentation and Reassembly" : "Fragmentation takes place in the transmitting PE immediately prior to PW encapsulation, and reassembly takes place in the receiving PE immediately after PW decapsulation.", in brief it uses the B and E bits in the control word as well as the sequence number, and since the control word is imposed on the ingress PE and then processed on the egress PE, thus the ingress PE would be capable of fragmentation and the egress PE would be capable of reassembly, again its all about the code you are running.
Now to the 1530 MTU point, again I'll take an AToM case (more specifically EoMPLS), the Core MTU must be equal or greater than the summation of: Edge MTU (1500:IP) + Transport Header (14:Ethr + 4:assuming dot1q) + Control Word (4) + MPLS Label Stack (4:LDP + 4:VC), which gives a total of 1530. I believe that the reason behind choosing the value of 1530 in the new codes lies in the fact that it is the most common highest MTU seen in practice, although a value of 1534 is required if EoMPLS VLAN + TE FRR is used, but anyway the MPLS MTU is user settable up to 1580.
I hope that I've been informative.
BR,
Mohammed Mahmoud.

MPLS MTU issues

Hello I am facing a strange issue and would be grateful to have any advice on this please.
I have an existing setup where Site A and site B are connected within an MPLS setup:
When I ping from Site B to site A with size 1500 it works no problem
ping vrf Data1 192.168.230.2 size 1500
Type escape sequence to abort.
Sending 5, 1500-byte ICMP Echos to 192.168.230.2, timeout is 2 seconds:
Success rate is 100 percent (5/5), round-trip min/avg/max = 496/496/500 ms
there is a new site C that has come up with a new link as a back up of site B but when i ping it only allows ping of 1496 and not anything more.
ping vrf Data1 192.168.230.2 size 1496
Type escape sequence to abort.
Sending 5, 1496-byte ICMP Echos to 192.168.230.2, timeout is 2 seconds:
Success rate is 100 percent (5/5), round-trip min/avg/max = 500/502/504 ms
ping vrf Data1 192.168.230.2 size 1497
Type escape sequence to abort.
Sending 5, 1497-byte ICMP Echos to 192.168.230.2, timeout is 2 seconds:
Success rate is 0 percent (0/5)
I have compared the configs and they are setup with same MPLS mtu size show as follows
Site B
interface FastEthernet2/0
no ip address
duplex full
speed 100
mpls label protocol ldp
mpls ip
mpls mtu 1524
end
Site C
interface FastEthernet0/1
no ip address
ip flow ingress
load-interval 30
duplex full
speed 100
mpls label protocol ldp
mpls ip
mpls mtu 1524
end
I have asked the provider of the circuit and they say that the MTU sizes on their devices are 1530 same as the link from site B because the switches connected to both the above routers have system MTU as 1530.
I cannot find out what the issue could be ?
any help on this will be much appreciated.
many thanks in advance.

Site A is 7206
Site B is 7206 and
Site C is 2811
1) Directions failing is both directions
When i debug the ip icmp echo
it comes back on Site A router as
*Oct 24 14:08:57.872 UTC: ICMP: dst (10.181.0.1) port unreachable rcv from 10.181.3.105
the utilization on Site C is as follows: could it be a reason why large sized packets are failing?
100 ** *       * ** * * *    * *      *   *         *        *   **     *
90 ** *       * ***** * *    * *      * *** *      * *      *   ****   *
80 #******    ** **###** **   ****      ****#*** *   ****     ** *##** **
70 #####***   ** *######********#**   * **####***#*****#**    ** *###***##
60 #######********###########*#####*******###############*** ****#########
50 ##########**#####################***####################******##########
40 ########################################################################
30 ########################################################################
20 ########################################################################
10 ########################################################################
   0....5....1....1....2....2....3....3....4....4....5....5....6....6....7..
             0    5    0    5    0    5    0    5    0    5    0    5    0
                   CPU% per hour (last 72 hours)
                  * = maximum CPU%   # = average CPU%
I have checked the configs and compared with the Site B configs and they seem to good. I have validated with someone else as well.
How do I check if there are places where more than one label is being used?

2801 MPLS MTU

2801 supports MPLS.
2801 does NOT support LAN Jumbo Frames (MTU > 1500)
MPLS requires > 1500 MTU
12.5 IOS will support > 1500 MTU on 2801
Has anyone run across this before? Interface command âip tcp adjust-mss 1460â is the only thing I could think of.
MPLS requires > 1500 MTU

Hi Tom
Try 12.4(20)T which supports MTU of 1600 on all ISR series.
http://shivlu.blogspot.com/2008/10/support-of-1600-mtu.html
regards
shivlu jain

MPLS MTU problem

Hey Guys,
We currently have several remote sites that connect back to our core (7206vxr) via p2p t1's. One Site connects back to us via MetroE. All Sites are setup under 1 vrf. The sites that have t1's can ping each other with no problem even bumping the datagram size up to over 1000. The problem I'm running into is with the MetroE site. If I try and ping it or ping from the site and bump the datagram size up to 250 or higher the success rate goes down to about 30%. This in turn seems to be causing some issues with our clients. Any advice on this would be great. The setup is all the interfaces that connect to the remote sites are in the same vrf. We then run EIGRP and route the subnets needed and everything works great besides a few applications that seem to be related to something with the MetroE,MPLS, and MTU size, I just not sure where to start.
Thanks in Advance

Ok I tried this 3 times and upped the DG size by 100 each time. At 300 I started getting packet loss. The thing I don't understand is when setting the "Set DF bit in IP header" to yes I don't notice any differnce than when it's set to No.
TX-OPT-RTR#ping
Protocol [ip]:
Target IP address: 192.168.0.254
Repeat count [5]: 10
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface:
Type of service [0]:
Set DF bit in IP header? [no]: y
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 10, 100-byte ICMP Echos to 192.168.0.254, timeout is 2 seconds:
Packet sent with the DF bit set
Success rate is 100 percent (10/10), round-trip min/avg/max = 32/32/32 ms
TX-OPT-RTR#ping
Protocol [ip]:
Target IP address: 192.168.0.254
Repeat count [5]: 10
Datagram size [100]: 200
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface:
Type of service [0]:
Set DF bit in IP header? [no]: y
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 10, 200-byte ICMP Echos to 192.168.0.254, timeout is 2 seconds:
Packet sent with the DF bit set
Success rate is 100 percent (10/10), round-trip min/avg/max = 32/32/36 ms
TX-OPT-RTR#ping
Protocol [ip]:
Target IP address: 192.168.0.254
Repeat count [5]: 10
Datagram size [100]: 300
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface:
Type of service [0]:
Set DF bit in IP header? [no]: y
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 10, 300-byte ICMP Echos to 192.168.0.254, timeout is 2 seconds:
Packet sent with the DF bit set
Success rate is 70 percent (7/10), round-trip min/avg/max = 32/33/36 ms

MPLS MTU

Similar Messages

Maybe you are looking for