10-10-2011 02:13 PM - edited 03-04-2019 01:52 PM
Hi All
Hope someone could shed some light on this
I have a cisco 2811 router that is runing a ipsec VPN between two of our offices and is connected to a 10MB line
The problem is it seems the router cant encrypt the data quicker enough for the line speed
with the ipsec on line max out at around 3mb/sec
with out and just the gre tunnel is does the full 10mb
As far as i know this router has a hardware chips for ipsec VPNs it list a VPN card under the show ver
is there something i have got wrong?
thanks
Ben
10-10-2011 02:39 PM
As far as i know this router has a hardware chips for ipsec VPNs it list a VPN card under the show ver
Not all. Try the command "sh crypto engine config". Look under "crypto engine type", if you have hardware then it's good. If you have "software" it could either mean that the VPN hardware was turned off or you don't have one.
Is CEF enabled?
Can you post the output for the command "sh interface tunnel"? What is your IOS and feature set?
10-10-2011 02:44 PM
Hi Leo
the engine type is hardware and cef is enabled
below is the output
hope that helps
thanks
Ben
Hardware is Tunnel
Description: VPN Tunnel to London
Internet address is 192.168.84.2/24
MTU 1514 bytes, BW 9 Kbit, DLY 500000 usec,
reliability 255/255, txload 245/255, rxload 231/255
Encapsulation TUNNEL, loopback not set
Keepalive not set
Tunnel source 1.1.1.1 (FastEthernet0/0), destination 2.2.2.2
tunnel protocol/transport GRE/IP
Key disabled, sequencing disabled
Checksumming of packets disabled
Tunnel TTL 255
Fast tunneling enabled
Tunnel transmit bandwidth 8000 (kbps)
Tunnel receive bandwidth 8000 (kbps)
Last input 00:00:00, output 00:00:00, output hang never
Last clearing of "show interface" counters never
Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 6
Queueing strategy: fifo
Output queue: 0/0 (size/max)
5 minute input rate 149000 bits/sec, 32 packets/sec
5 minute output rate 49000 bits/sec, 31 packets/sec
110718 packets input, 85718754 bytes, 0 no buffer
Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored, 0 abort
93037 packets output, 15133848 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 output buffer failures, 0 output buffers swapped out
10-10-2011 04:11 PM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
Is your VPN configuration defined to avoid fragmentation?
10-11-2011 12:51 AM
Hi Joseph
I am not sure is there some way i can check?
thanks
Ben
10-11-2011 02:12 AM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
Review: http://www.cisco.com/en/US/tech/tk827/tk369/technologies_white_paper09186a00800d6979.shtml
10-11-2011 06:03 AM
Hi Ben,
Just a note to what Joseph mentioned. You can check if the router is fragmenting/reassembling via the command: show ip traffic.
From the show interface output, we have the default MTU so I would not be surprised if we indeed see some fragmentation or reassembly. We cannot reassemble the packets in CEF so we have to punt them to process switching.
It might be also worth checking if there is any other CPU intensive feature configured on the tunnel (eg. QoS, NBAR).
Warm Regards,
Rose
10-11-2011 06:08 AM
Hi Rose
we do have NBAR running but it is not showing as high on the CPU useage but will take it of for now
is there a recommened MTU size for GRE tunnels? if the packet is to large for the MTU is there anyway around it?
thanks
Ben
10-11-2011 06:21 AM
Hello Ben,
A quite general MTU value for GRE tunnels is: 1400. You might want to add the tcp adjust-mss value: 1360 though it only affects TCP traffic. These should be configured on both endpoints of the tunnel.
What is important that we fragment the packet before it is encrypted. If we fragment it after encryption, the remote end has to buffer the fragments, reassemble the packets before it could encrypt them and this adds overhead to the CPU. If we set the MTU, we will fragment the packets before they are encrypted so the reassembly does not have to happen on the receiving router.
Warm Regards,
Rose
10-11-2011 06:50 AM
Hello Rose
I have set both ends of the tunnel to 1400 using "ip mtu 1400" but has made no difference
the CPU is running at around 77% when coping a file over the tunnel
should i try al ower MTU?
many thanks
Ben
10-11-2011 06:57 AM
Hello Ben,
What you do see if you check the show ip traffic (several times during the file transfer and on both sides)? Do you see any fragmens/reassembly? If not, does not make sense to try tuning the MTU more.
But we could check if there is any other feature that might require CPU cycles and we could check what is the average packet size (show cry eng acc stat and show int tun should help here).
Warm Regards,
Rose
10-11-2011 07:02 AM
Hello Rose
they are and i am also getting
16:44:45: %IP_VFR-4-FRAG_TABLE_OVERFLOW: FastEthernet0/0: the fragment table has reached its maximum
threshold 16
have had quick google and can up the limit but guess there a problem with the mtu somewhere in the first place that causing the problem?
thanks
Ben
10-11-2011 07:23 AM
Hi Ben,
Yes, the router received too many fragments and it could not handle them.
We can try is to ping across the tunnel with the df-bit set. This would help us to determine the maximum MTU on the tunnel itself and see if it is causing the fragmentation. If it is less than 1400, we will need to decrease it further or to find the device which limits the MTU.
Warm Regards,
Rose
10-11-2011 07:30 AM
Hi Rose
have done a ping with the MTU of 1400 and the DF set and it went through fine
is the problem when the GRE packet gets encapsulated in the ipsec header it is going over the MTU of the outbound WAN interface?
thanks
Ben
10-11-2011 07:36 AM
Would a good test be to try it with the ipsec and not a gre tunnel and see what the throughput is?
thanks
Ben
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide