02-16-2017 11:31 AM - edited 03-05-2019 08:02 AM
Hi Guys,
Hoping someone can help
we have an satellite office connected back to us over 2 Metro Ethernet links 10Mb & 20Mb, for where we present internet transit.
I have 4 cisco 2901s on each side of the links and tried using GLBP, to fully utilise bandwidth & provide failover.
This works to a fashion, however using this method client will only use one virtual forwarder hence never achieve the full combined bandwidth of 30Mb.
is there anything i can use to actually utilise the links as one, such as port channel which obviously i cant use due to unequal speed links. I need the full 30Mb especially for the veeam replication from site.
Thanks in advance
Solved! Go to Solution.
02-16-2017 12:51 PM
The only thing that comes to my mind, to effectively use both links unequal bandwidths as one logical aggregate, would be some kind of hardware mux at both ends. (Don't know if any such thing even exists.)
If you don't need any individual flow to use all 30 Mbps, your next best option would be something like PfR that can dynamically load balance multiple flows across both links.
02-16-2017 11:47 AM
Hello,
Can you provide me the GLBP configuration? Which load-balance algorithm are you using?
Regards,
02-16-2017 12:25 PM
Hi anaid_30kadi,
I'm trying weghted.
logic being 40 weight to 20 weight should give me 2to1?
config as below RO-01 both end on the 10MB, RO-02 on the 20Mb
Head Office
HQ RO-01
track 1 ip route 1.1.1.1 255.255.255.255 reachability
interface Loopback0
ip address 3.3.3.3 255.255.255.0
!
interface Embedded-Service-Engine0/0
no ip address
shutdown
!
interface GigabitEthernet0/0
description $ETH-LAN$$ETH-SW-LAUNCH$$INTF-INFO-GE 0/0$
ip address 172.17.17.1 255.255.255.0
glbp 1 ip 172.17.17.254
glbp 1 priority 101
glbp 1 preempt
glbp 1 weighting 20 lower 10
glbp 1 load-balancing weighted
glbp 1 weighting track 1 decrement 10
duplex auto
speed auto
!
interface GigabitEthernet0/1
ip address 172.16.17.2 255.255.255.252
duplex full
speed 100
!
ip forward-protocol nd
ip route 0.0.0.0 0.0.0.0 172.17.17.250
ip route 1.1.1.1 255.255.255.255 172.16.17.1
ip route 10.0.100.0 255.255.255.0 172.16.17.1
HQ-RO-02
track 1 ip route 4.4.4.4 255.255.255.255 reachability
!
interface Loopback0
ip address 2.2.2.2 255.255.255.255
!
interface Embedded-Service-Engine0/0
no ip address
shutdown
!
interface GigabitEthernet0/0
description $ETH-LAN$$ETH-SW-LAUNCH$$INTF-INFO-GE 0/0$
ip address 172.17.17.2 255.255.255.0
glbp 1 ip 172.17.17.254
glbp 1 preempt
glbp 1 weighting 40 lower 20
glbp 1 load-balancing weighted
glbp 1 weighting track 1 decrement 20
duplex auto
speed auto
!
interface GigabitEthernet0/1
ip address 172.16.17.6 255.255.255.252
duplex full
speed 100
!
ip forward-protocol nd
!
ip route 0.0.0.0 0.0.0.0 172.17.17.250
ip route 4.4.4.4 255.255.255.255 172.16.17.5
ip route 10.0.100.0 255.255.255.0 172.16.17.5
Satellite Office
SAT-RO-01
track 1 ip route 3.3.3.3 255.255.255.255 reachability
!
interface Loopback0
ip address 1.1.1.1 255.255.255.255
!
interface Embedded-Service-Engine0/0
no ip address
shutdown
!
interface GigabitEthernet0/0
ip address 10.0.100.252 255.255.255.0
glbp 1 ip 10.0.100.254
glbp 1 priority 101
glbp 1 preempt
glbp 1 weighting 20 lower 10
glbp 1 load-balancing weighted
glbp 1 weighting track 1 decrement 20
duplex auto
speed auto
!
interface GigabitEthernet0/1
ip address 172.16.17.1 255.255.255.252
duplex auto
speed auto
!
ip forward-protocol nd
!
ip route 0.0.0.0 0.0.0.0 172.16.17.2
ip route 3.3.3.3 255.255.255.255 172.16.17.2
SAT-RO-02
track 1 ip route 2.2.2.2 255.255.255.255 reachability
!
!
!
!
!
interface Loopback0
ip address 4.4.4.4 255.255.255.255
!
interface Embedded-Service-Engine0/0
no ip address
shutdown
!
interface GigabitEthernet0/0
description $ETH-LAN$$ETH-SW-LAUNCH$$INTF-INFO-GE 0/0$
ip address 10.0.100.253 255.255.255.0
glbp 1 ip 10.0.100.254
glbp 1 preempt
glbp 1 weighting 40 lower 20
glbp 1 load-balancing weighted
glbp 1 weighting track 1 decrement 20
duplex auto
speed auto
!
interface GigabitEthernet0/1
ip address 172.16.17.5 255.255.255.252
duplex auto
speed auto
!
ip forward-protocol nd
!
ip route 0.0.0.0 0.0.0.0 172.16.17.6
ip route 2.2.2.2 255.255.255.255 172.16.17.6
Thanks for your time to look into this, this is the first time i'm using GLBP :-(
sh GLBP on HQ-RO-01 head office end
GigabitEthernet0/0 - Group 1
State is Active
4 state changes, last state change 07:06:28
Virtual IP address is 172.17.17.254
Hello time 3 sec, hold time 10 sec
Next hello sent in 1.088 secs
Redirect time 600 sec, forwarder timeout 14400 sec
Preemption enabled, min delay 0 sec
Active is local
Standby is 172.17.17.2, priority 100 (expires in 7.520 sec)
Priority 101 (configured)
Weighting 20 (configured 20), thresholds: lower 10, upper 20
Track object 1 state Up decrement 10
Load balancing: weighted
Group members:
00a2.eef2.daf0 (172.17.17.1) local
00a2.eef2.dd30 (172.17.17.2)
There are 2 forwarders (1 active)
Forwarder 1
State is Listen
MAC address is 0007.b400.0101 (learnt)
Owner ID is 00a2.eef2.dd30
Redirection enabled, 597.536 sec remaining (maximum 600 sec)
Time to live: 14397.536 sec (maximum 14400 sec)
Preemption enabled, min delay 30 sec
Active is 172.17.17.2 (primary), weighting 40 (expires in 9.440 sec)
Client selection count: 297
Forwarder 2
State is Active
3 state changes, last state change 07:06:06
MAC address is 0007.b400.0102 (default)
Owner ID is 00a2.eef2.daf0
Redirection enabled
Preemption enabled, min delay 30 sec
Active is local, weighting 20
Client selection count: 275
sh GLBP on SAT0-RO-01 Satellite office end
GigabitEthernet0/0 - Group 1
State is Active
1 state change, last state change 07:19:29
Virtual IP address is 10.0.100.254
Hello time 3 sec, hold time 10 sec
Next hello sent in 1.440 secs
Redirect time 600 sec, forwarder timeout 14400 sec
Preemption enabled, min delay 0 sec
Active is local
Standby is 10.0.100.253, priority 100 (expires in 9.856 sec)
Priority 101 (configured)
Weighting 20 (configured 20), thresholds: lower 10, upper 20
Track object 1 state Up decrement 20
Load balancing: weighted
Group members:
00a2.eef2.e318 (10.0.100.253)
00eb.d565.f5ee (10.0.100.252) local
There are 2 forwarders (1 active)
Forwarder 1
State is Listen
6 state changes, last state change 03:36:57
MAC address is 0007.b400.0101 (learnt)
Owner ID is 00a2.eef2.e318
Redirection enabled, 599.872 sec remaining (maximum 600 sec)
Time to live: 14399.872 sec (maximum 14400 sec)
Preemption enabled, min delay 30 sec
Active is 10.0.100.253 (primary), weighting 40 (expires in 11.424 sec)
Client selection count: 16055
Forwarder 2
State is Active
3 state changes, last state change 00:18:46
MAC address is 0007.b400.0102 (default)
Owner ID is 00eb.d565.f5ee
Redirection enabled
Preemption enabled, min delay 30 sec
Active is local, weighting 20
Client selection count: 3249
Thanks again any guidance is greatly appreciated.
02-16-2017 03:21 PM
Hello,
It should work if you have a good traffic distribution. Make sure GLBP members see each other through layer 2. GLBP would cause asymmetrical routing. Consider this if you have a security device in the network.
Masous
02-16-2017 12:51 PM
The only thing that comes to my mind, to effectively use both links unequal bandwidths as one logical aggregate, would be some kind of hardware mux at both ends. (Don't know if any such thing even exists.)
If you don't need any individual flow to use all 30 Mbps, your next best option would be something like PfR that can dynamically load balance multiple flows across both links.
02-16-2017 01:50 PM
Thanks Joseph,
GLBP wont do what i need, learned the hard way :-(
Now im thinking to just to use 2 HSRP groups each end (For failover) then use PBR to route traffic by class / port.
I'm thinking PBR on RO-01 with tracking each end,
If the 20MB link fails (RO-02 either end) tracking on RO-01 should take care of it
If the 10Mb link drops again the PBR tracking should accommodate
if either end RO-01 fails HSRP wil jump to RO-02 which has no PBR.
Lets see :-)
thanks for your advice, if you see anything wrong in my logic above let me know
02-17-2017 06:36 AM
Yea, you could use PBR, but it's a pain to manage. Link fail-over can be a bit tricky with it too.
Perhaps a better way for "static" load balancing, over GLBP, would be to use routing load balancing. You might use a routing protocol that supports ECMP (just "lie" to the routing protocol), or if using EIGRP, use its support for unECMP.
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