08-20-2016 06:48 AM - edited 03-05-2019 04:32 AM
To sum up: 2911 dmvpn tunnel egress throughput increased slightly and there is less cpu usage after activating the ISM-VPN card, but dmvpn tunnel ingress throughput actually decreased and there is more cpu usage. The expectation was that both ingress and egress throughput would have increased substantially after activating the card based on Cisco documentation, any suggestions are appreciated.
Expectation for tunneled ipsec traffic without ISM-VPN card installed -
• The SEC-K9 license limits encrypted throughput to less than or equal to 85-Mbps unidirectional traffic in or out of the ISR G2 router, with a bidirectional total of 170 Mbps. This requirement applies for the Cisco 1900, 2900, and 3900 ISR G2 platforms.
Expectation for tunneled ipsec traffic with ISM-VPN card installed, low IMIX number being the real world number:
• The Cisco 2900 Series Module (ISM-VPN-29) can provide hardware-based IPSec encryption services of 145 and 550 Mbps in the Cisco 2901, 150 and 600 Mbps in the Cisco 2911, 220 and 700 Mbps in the Cisco 2921, and 385 and 900 Mbps in the Cisco 2951 (IPSec IMIX and 1400-byte packets).
spoke is the 2911 + ISM-VPN card being tested
hub is an ASR1002 nat'd behind an ASA5520, no changes are being made at the hub during these tests.
dmvpn throughput with no Crypto engine slot 0 or before adding the card = ~50Mb up from spoke to hub, ~25-30Mb down from hub to spoke
dmvpn throughput with Crypto engine slot 0 after adding the card = ~65-70Mb up from spoke to hub, ~15-20Mb down from hub to spoke
numbers came up similar in repeat tests with iperf, smb, ftp transfers over an extended period
spoke isp bandwidth 200Mb/200Mb, internet non-tunnel speed test from lan test machine = 125Mb down / 125Mb up (2911 performance limitation)
hub isp bandwidth 500Mb/500Mb, internet non-tunnel speed test from lan test machine = 450Mb down / 375Mb up
unencrypted peering throughput test between these locations is still pending.
HUB -
NAME: "Chassis", DESCR: "Cisco ASR1002 Chassis" PID: ASR1002
NAME: "module F0", DESCR: "Cisco ASR1000 Embedded Services Processor, 5Gbps" PID: ASR1000-ESP5 , VID: V03,
NAME: "module 0", DESCR: "Cisco ASR1002 SPA Interface Processor 10" PID: ASR1002-SIP10
NAME: "subslot 0/0 transceiver 0", DESCR: "GE T" PID: SP7041-E , VID: E
NAME: "module R0", DESCR: "Cisco ASR1002 Route Processor 1" PID: ASR1002-RP1 , VID: V05,
sh ver hub - asr1000rp1-adventerprisek9.03.13.05.S.154-3.S5-ext.bin
hub -
crypto ipsec security-association replay window-size 1024
! Warning: window size of 512 actually used
!
crypto ipsec transform-set DMVPNtrans esp-des esp-md5-hmac
mode transport
crypto ipsec df-bit clear
!
interface Tunnel0
ip address 192.168.0.1 255.255.255.0
no ip redirects
ip mtu 1400
no ip split-horizon eigrp 100
no ip next-hop-self eigrp ***
ip nhrp authentication ***
ip nhrp map multicast dynamic
ip nhrp network-id ***
ip nhrp holdtime 300
ip nhrp max-send 200 every 10
ip nhrp redirect
ip tcp adjust-mss 1360
load-interval 30
delay 10000000
qos pre-classify
keepalive 10 3
cdp enable
tunnel source Loopback0
tunnel mode gre multipoint
tunnel key ****
tunnel bandwidth transmit 5000000
tunnel bandwidth receive 5000000
tunnel protection ipsec profile ***
!
interface Loopback0
description hub receives ASA public internet nat here
ip address 192.168.248.1 255.255.255.240
no ip redirects
no ip unreachables
ip mtu 1400
!
interface GigabitEthernet0/0/0
description hub lan
ip address 192.168.0.210 255.255.252.0
no ip redirects
no ip unreachables
ip access-group external-security in
negotiation auto
!
everything below is from the spoke -
sh inv -
NAME: "CISCO2911/K9", DESCR: "CISCO2911/K9 chassis, Hw Serial#: , Hw Revision: 1.0"
NAME: "Internal Services Module - Crypto Engine on Slot 0", DESCR: "Internal Services Module - Crypto Engine" PID: ISM-VPN-29 , VID: V02 , SN:
sh ver - running c2900-universalk9-mz.SPA.152-4.M1.bin
sh lic -
Index 9 Feature: hseck9
Period left: Life time
License Type: Permanent
License State: Active, In Use
License Count: Non-Counted
License Priority: Medium
Index 2 Feature: securityk9
Period left: Life time
License Type: Permanent
License State: Active, In Use
License Count: Non-Counted
License Priority: Medium
sh log -
*Aug 20 13:34:09.507: %VPN_HW-6-INFO_LOC: Crypto engine: onboard 0 State changed to: Initialized
*Aug 20 13:34:09.511: %VPN_HW-6-INFO_LOC: Crypto engine: onboard 0 State changed to: Enabled
*Aug 20 2016 06:34:16.203 MST: %VPN_HW-6-INFO_LOC: Crypto engine: slot 0 State changed to: Initialized
*Aug 20 2016 06:34:16.203 MST: %VPN_HW-6-INFO_LOC: Crypto engine: onboard 0 State changed to: Disabled
*Aug 20 2016 06:34:16.223 MST: %VPN_HW-6-INFO_LOC: Crypto engine: slot 0 State changed to: Enabled
*Aug 20 2016 06:34:16.223 MST: ISM VPN UP & READY
sh diag -
Internal Services Module (ISM) Slot 0
Internal Services Module - Crypto Engine
Internal Services Module is analyzed
Internal Services Module insertion time 00:04:05 ago
EEPROM contents at hardware discovery:
PCB Serial Number : ***
Hardware Revision : 1.0
Part Number : 73-13342-05
Top Assy. Part Number : 800-37237-02
Board Revision : A0
Deviation Number : 0
Fab Version : 02
Product (FRU) Number : ISM-VPN-29
Version Identifier : V02
CLEI Code : CMUCACLBAA
sh cry eng conf -
crypto engine name: Virtual Private Network (VPN) Module
crypto engine type: hardware
State: Enabled
Location: slot 0
Product Name: ISM VPN Accelerator
UBOOT Ver : U-Boot 1.1.1 - ISRG2-Crypto-Engine - Version 2.7 (Build time: Mar 7 2011 - 09:12:23)
Firmware Ver: User: dranjit - View/Label: REVENTON_FW_COMMIT_IOS_07122012 - Date: Jul 12 2012 - Time: 18:28:58
HW State : READY
sh run -
crypto ipsec transform-set DMVPNtrans esp-des esp-md5-hmac
mode transport
!
crypto ipsec profile ***
set transform-set DMVPNtrans
interface Tunnel0
description DMVPN
bandwidth 3000
ip address ***
no ip redirects
ip mtu 1400
ip flow ingress
ip flow egress
ip nat inside
ip nhrp authentication ***
ip nhrp map multicast dynamic
ip nhrp map multicast ***
ip nhrp map ***
ip nhrp network-id ***
ip nhrp holdtime 300
ip nhrp nhs ***
ip virtual-reassembly in
ip tcp adjust-mss 1360
load-interval 30
tunnel source GigabitEthernet0/2
tunnel mode gre multipoint
tunnel key ***
tunnel protection ipsec profile ***
interface GigabitEthernet0/2
description spoke isp
bandwidth 200000
ip address 2.2.2.2 255.255.255.240
ip access-group edge-ingress in
ip access-group edge-egress out
ip flow ingress
ip flow egress
ip nat outside
ip virtual-reassembly in
duplex full
speed 1000
no cdp enable
!
interface GigabitEthernet0/0
description spoke lan
ip address 192.168.4.1 255.255.252.0
ip flow ingress
ip flow egress
ip nat inside
ip virtual-reassembly in
ip policy route-map path-control
duplex auto
speed auto
!
download from hub to spoke before activating ISM-VPN card -
iperf -
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 3.02 MBytes 25.3 Mbits/sec
[ 4] 1.00-2.01 sec 3.50 MBytes 29.2 Mbits/sec
[ 4] 2.01-3.00 sec 3.26 MBytes 27.5 Mbits/sec
[ 4] 3.00-4.00 sec 3.20 MBytes 26.8 Mbits/sec
[ 4] 4.00-5.00 sec 3.38 MBytes 28.4 Mbits/sec
[ 4] 5.00-6.00 sec 3.24 MBytes 27.2 Mbits/sec
[ 4] 6.00-7.00 sec 3.40 MBytes 28.5 Mbits/sec
[ 4] 7.00-8.00 sec 3.39 MBytes 28.4 Mbits/sec
[ 4] 8.00-9.00 sec 3.38 MBytes 28.4 Mbits/sec
[ 4] 9.00-10.00 sec 3.28 MBytes 27.5 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 33.2 MBytes 27.8 Mbits/sec sender
[ 4] 0.00-10.00 sec 33.2 MBytes 27.8 Mbits/sec receiver
CPU Utilization: local/receiver 5.6% (2.1%u/3.5%s), remote/sender 0.3% (0.1%u/0.1%s)
2911 cpu, each peak is a test -
5555544444 44444555551111111111444445555533333
911111333334444477777444441111188888222224444466666000003333
100
90
80
70
60 *****
50 ***** ***** *****
40 ********** ********** **********
30 ********** ********** ***************
20 ********** *************** ***************
10 *********** ****************************************
0....5....1....1....2....2....3....3....4....4....5....5....6
0 5 0 5 0 5 0 5 0 5 0
CPU% per second (last 60 seconds)
download from hub to spoke after activating ISM-VPN card -
iperf
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.01 sec 1.53 MBytes 12.7 Mbits/sec
[ 4] 1.01-2.00 sec 1.72 MBytes 14.5 Mbits/sec
[ 4] 2.00-3.01 sec 809 KBytes 6.58 Mbits/sec
[ 4] 3.01-4.01 sec 550 KBytes 4.51 Mbits/sec
[ 4] 4.01-5.00 sec 1.67 MBytes 14.1 Mbits/sec
[ 4] 5.00-6.00 sec 1.99 MBytes 16.7 Mbits/sec
[ 4] 6.00-7.01 sec 1.86 MBytes 15.4 Mbits/sec
[ 4] 7.01-8.00 sec 1.89 MBytes 16.0 Mbits/sec
[ 4] 8.00-9.00 sec 2.28 MBytes 19.2 Mbits/sec
[ 4] 9.00-10.01 sec 1.97 MBytes 16.5 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.01 sec 16.3 MBytes 13.7 Mbits/sec sender
[ 4] 0.00-10.01 sec 16.3 MBytes 13.7 Mbits/sec receiver
CPU Utilization: local/receiver 3.0% (1.6%u/1.4%s), remote/sender 0.0% (0.0%u/0.0%s)
each cpu peak is a throughput test -
99999999999999999999999 222229999977777 9999999
444444446666699999999998888844444777774444433333777773333300
100 *************** *****
90 *********************** ***** *******
80 *********************** ***** *******
70 *********************** ********** *******
60 *********************** ********** *******
50 *********************** ********** *******
40 *********************** ********** *******
30 *********************** ********** *******
20 *********************** *************** *******
10 ******************************************* ************
0....5....1....1....2....2....3....3....4....4....5....5....6
0 5 0 5 0 5 0 5 0 5 0
CPU% per second (last 60 seconds)
upload from spoke to hub before activating ISM-VPN card -
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 5.30 MBytes 44.4 Mbits/sec
[ 4] 1.00-2.00 sec 5.53 MBytes 46.4 Mbits/sec
[ 4] 2.00-3.00 sec 5.66 MBytes 47.5 Mbits/sec
[ 4] 3.00-4.00 sec 5.92 MBytes 49.7 Mbits/sec
[ 4] 4.00-5.00 sec 5.68 MBytes 47.6 Mbits/sec
[ 4] 5.00-6.00 sec 5.89 MBytes 49.4 Mbits/sec
[ 4] 6.00-7.00 sec 6.01 MBytes 50.4 Mbits/sec
[ 4] 7.00-8.00 sec 5.72 MBytes 48.0 Mbits/sec
[ 4] 8.00-9.00 sec 5.88 MBytes 49.4 Mbits/sec
[ 4] 9.00-10.00 sec 6.05 MBytes 50.8 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 57.9 MBytes 48.6 Mbits/sec sender
[ 4] 0.00-10.00 sec 57.9 MBytes 48.6 Mbits/sec receiver
CPU Utilization: local/receiver 10.7% (4.3%u/6.4%s), remote/sender 0.8% (0.3%u/0.5%s)
upload from spoke to hub after activating ISM-VPN card:
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 8.00 MBytes 67.0 Mbits/sec
[ 4] 1.00-2.01 sec 8.00 MBytes 66.8 Mbits/sec
[ 4] 2.01-3.01 sec 8.00 MBytes 67.1 Mbits/sec
[ 4] 3.01-4.01 sec 8.00 MBytes 67.1 Mbits/sec
[ 4] 4.01-5.00 sec 9.00 MBytes 75.8 Mbits/sec
[ 4] 5.00-6.01 sec 8.00 MBytes 66.8 Mbits/sec
[ 4] 6.01-7.01 sec 8.00 MBytes 67.1 Mbits/sec
[ 4] 7.01-8.01 sec 8.00 MBytes 67.1 Mbits/sec
[ 4] 8.01-9.01 sec 8.00 MBytes 67.1 Mbits/sec
[ 4] 9.01-10.01 sec 9.00 MBytes 75.5 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.01 sec 82.0 MBytes 68.7 Mbits/sec sender
[ 4] 0.00-10.01 sec 81.1 MBytes 67.9 Mbits/sec receiver
CPU Utilization: local/sender 0.9% (0.3%u/0.6%s), remote/receiver 1.5% (0.5%u/1.0%s)
99999999933333
555577777999993333355555444444444433333444445555533333333333
100 *********
90 *********
80 *********
70 *********
60 *********
50 *********
40 **************
30 **************
20 **************
10 ************** ***** *****
0....5....1....1....2....2....3....3....4....4....5....5....6
0 5 0 5 0 5 0 5 0 5 0
CPU% per second (last 60 seconds)
only "IP Input" shows up using much cpu in the processes while pushing a test transfer:
upload from spoke to hub with ism-vpn enabled -
sh proc cpu sorted
CPU utilization for five seconds: 97%/72%; one minute: 40%; five minutes: 13%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
168 56308 137900 408 22.79% 8.92% 2.57% 0 IP Input
download from hub to spoke with ism-vpn enabled -
CPU utilization for five seconds: 84%/40%; one minute: 65%; five minutes: 29%
PID Runtime(ms) Invoked uSecs 5Sec 1Min 5Min TTY Process
168 99572 160143 621 40.23% 19.40% 7.32% 0 IP Input
also will note that with or without ism-vpn enabled there is slight improvement in throughput if ip policy route-map path-control which we use for 3 internet traffic routes unrelated to dmvpn is removed from spoke lan interface or ACL removed from internet interface, ingress spoke to hub throughput still remains poor. it's understood that 2911 aggregate throughout performance will take a hit with acl & route-maps in use -
egress spoke to hub:
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.01 sec 9.00 MBytes 75.0 Mbits/sec
[ 4] 1.01-2.01 sec 10.0 MBytes 83.9 Mbits/sec
[ 4] 2.01-3.01 sec 10.0 MBytes 83.8 Mbits/sec
[ 4] 3.01-4.01 sec 10.0 MBytes 83.9 Mbits/sec
[ 4] 4.01-5.01 sec 10.0 MBytes 83.9 Mbits/sec
[ 4] 5.01-6.01 sec 2.00 MBytes 16.8 Mbits/sec
[ 4] 6.01-7.01 sec 7.00 MBytes 58.7 Mbits/sec
[ 4] 7.01-8.01 sec 9.00 MBytes 75.5 Mbits/sec
[ 4] 8.01-9.01 sec 10.0 MBytes 83.9 Mbits/sec
[ 4] 9.01-10.01 sec 10.0 MBytes 83.8 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.01 sec 87.0 MBytes 72.9 Mbits/sec sender
[ 4] 0.00-10.01 sec 86.6 MBytes 72.6 Mbits/sec receiver
CPU Utilization: local/sender 1.1% (0.1%u/0.9%s), remote/receiver 2.3% (0.6%u/1.7%s)
ingress hub to spoke:
Starting Test: protocol: TCP, 1 streams, 1048576 byte blocks, omitting 0 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.01 sec 1.52 MBytes 12.7 Mbits/sec
[ 4] 1.01-2.00 sec 1.54 MBytes 13.0 Mbits/sec
[ 4] 2.00-3.00 sec 1.71 MBytes 14.4 Mbits/sec
[ 4] 3.00-4.00 sec 1.94 MBytes 16.2 Mbits/sec
[ 4] 4.00-5.00 sec 1.82 MBytes 15.3 Mbits/sec
[ 4] 5.00-6.00 sec 2.51 MBytes 21.0 Mbits/sec
[ 4] 6.00-7.01 sec 1.96 MBytes 16.3 Mbits/sec
[ 4] 7.01-8.01 sec 1.95 MBytes 16.4 Mbits/sec
[ 4] 8.01-9.00 sec 2.12 MBytes 17.9 Mbits/sec
[ 4] 9.00-10.00 sec 1.98 MBytes 16.5 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 19.1 MBytes 16.0 Mbits/sec sender
[ 4] 0.00-10.00 sec 19.1 MBytes 16.0 Mbits/sec receiver
CPU Utilization: local/receiver 4.4% (2.6%u/1.8%s), remote/sender 0.3% (0.1%u/0.2%s)
08-23-2016 06:50 AM
http://www.cisco.com/c/en/us/products/collateral/routers/3900-series-integrated-services-routers-isr/q-and-a-c67-606268.html
01-06-2017 09:18 AM
Just checking back to see if there are any suggestions or experience with these before I toss it and move on to looking at other hardware. 2911's severe throughput limitations with acl + route maps in use are a known factor but performance overall actually seems worse since this card was enabled. Example - high latency occurs on pings from lan client to lan interface or ping from locally within router and is reflected in response of network resources sporadically when pushing SMB traffic across the dmvpn link at 10-25Mb with corresponding cpu spike until traffic subsides.
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: