ECMP - equal cost multi path

Hi,
What is the concept behind ECMP (equal cost multi path) ? Is it different for EIGRP , OSPF , ISIS etc ?
thanks

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
The concept behind ECMP is to actively and concurrently take advantage of multiple link/path bandwidth.
Oh, and to just add to what Jon has already noted, ECMP usually doesn't track actual load. So, "seeing" unequal path/link loading, especially short term, isn't unusual.
PS:
There's also unequal cost multi-path routing too - EIGRP supports that.

Similar Messages

Equal Cost Multi Path route support.

Can anyone tell me if Solaris 8 supports ECMP? First if it works, and second, if it officially works?
Anyone have a web page or document stating it works?
Even more generally, can anyone point me to a Sun web page that lists RFC compliance for their OS's?
Thanks in advance!

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
The concept behind ECMP is to actively and concurrently take advantage of multiple link/path bandwidth.
Oh, and to just add to what Jon has already noted, ECMP usually doesn't track actual load. So, "seeing" unequal path/link loading, especially short term, isn't unusual.
PS:
There's also unequal cost multi-path routing too - EIGRP supports that.

EIGRP- Equal Cost Paths Selection

Hello,
This might be a simple question; using EIGRP and having 2 equal cost paths to the destination; how does the router/L3 switch route select the path from those two equal cost paths? Is there a way to force one path over anthoer?
Thanks in advance.
Best, ~sK

To answer your question we need to be clear that there are 2 parts to the decision about which path will be used to forward traffic. The first part is the identification of paths and their insertion into the routing table. This is the part where EIGRP plays a role. If EIGRP determines that there are two equal cost viable paths then both of the paths will be put into the routing table. This ends EIGRP involvement in the process. The part of the process that takes a particular packet and determines which path to use is handled by CEF. By default CEF looks at the number of available paths toward a destination and when there are multiple paths CEF will use a process that calculates based on source and destination address to choose a particular path. For the same source address and same destination address CEF should choose the same path every time.
HTH
Rick

RIP/EIGRP maximum number of Equal cost routes

I am looking for documentation on:
If a router has 20 equal cost routing paths and only 4 show up in the routing table, what is the determining factor that chooses which four and where are the other 16 put in case the 4 routes become unusable.
Most documentation will tell you the maximum amount and how to change it but I have not found any documentation on how EIGRP or RIP chooses the routes.

You can have up to 4 by default, but you can configure the routing table to accept up to 6 on older code, and up to 8 on newer code, using the command "maximum-paths" under router eigrp. EIGRP will attempt to install all of the available paths, but the routing table will only allow it to install the first x that it installs.
The routes which are not installed by EIGRP are placed in a "backup table." If one route fails, each routing protocol running on the router is notified, and will attempt to install any routes (which match the destination, of course) back into the routing table. Again, the frst x paths would win.
Be careful with this much redundancy in eigrp--you're playing with fire if you have 15 or 20 links between a pair of routers.Russ

IS-IS and IPv6 in equal cost path topology

Network of 7609s running 12.2(18)SXE3. Two
equal cost path links between each 7609. IS-IS Level 2 routers only. When pinging an off campus IPv6 host every other packet is lost. It seems like the equal cost paths prevent an alternative path off campus but that path isn't working.
We cannot figure out where to look next.
Any thoughts ? Is this possibly an IS-IS
issue ?
Thanks

OK. I'll try. We have 5-7609s running 12.2(18)SXE3.
Each 7609 has a variety of 100/1000Base Ethernet
interfaces. They each also have one WS-X6704-10GE out
of which connections to our primary and secondary
backbones originate. This is how the routers are
interconnected.
Three of the five core routers are currently running
IS-IS. If we ssh into any router and ping
www.netbsd.org every other packet is lost.
At the end of the network are two 6506 switches each
with a 10GE module and a 1GE module.
Attached to the 1GE module are fiber links two
two different Juniper M10s which also run IS-IS
and through which we get upstream connectivity
to our provider.
All of our routers are level 2 routers in the same area.
Does this help? If not please let me know what else
you would need.
Thanks,Steve

Nexus 7010 OSPF Equal Cost Paths

Hello,
I currently have two physical links connecting one data center to another. These are both 10Gb links and I have manually set the cost to the primary link to '1' and the secondary link to '10'. My question is, if I set the secondary link to '1' they would have equal cost routes. What is the selection process at this point? Will equal cost load balancing automatically kick in and use both links?
Thank you,
Scott

Scott
Haven't used Nexus switches but generally yes it should do depending on the routing protocol ie. statics, EIGRP, OSPF etc. will use equal cost paths if they are in the routing table.
BGP is different in that without further configuration it picks just one path so there is only one entry in the routing table.
By default it will use per destination load sharing and the default on Nexus is destination IP address and port number to choose which link to use.
"sh ip load-sharing"
will show you the current method it is using.
Jon

OSPF equal cost path

Hi all,
I have
The network 192.168.0.0 is directly connected to Router A and B. Router A and B are redistributing the network in the ospf area 0.
Router A,B,C,D belong to the same OSPF Area.
Router C is DR. The cost of the link is indicated in the draw.
Question:
Router D (and router C) is using as next hop the Router A for reaching the network 192.168.0.0.
Why the router D is not using the router B as next hop due to the fact that all the possible paths are equal cost?
Thanks.

Hi,
At the end the information provided by Rolf fit in my case.
Playing with the cost to reach the ASBR and the metric of the external route I was able to figure out why the router A is the50
preferred exit point: during the redistribution router A is modifying the metric value, in this way C and D are using A as
Thanks all!!!

2 locations, 2 core switch stacks, fibre in between, equal cost load balancing between?

Hi,
We've recently inherited a job that another company was doing, so we've had our hand slightly forced on the kit and overall topology involved, however that's all fine and we can make it work.
This is a collapsed core topology with core and access switches, split over 3 blocks (fibre connections between), one core switch/stack is in block B and the other in block C, with access switches throughout.
They require all access switches to be connected to the Core in B and the Core in C, and then obviously cross connects between the two cores.
They state:
"Core switches shall be linked with 2x 1Gbps links bonded into a standard compliant Etherchannel"
"Uplinks between access and core switches shall be non-blocking - for example equal cost load balancing at layer 3, or layer 2 bonded multi-chassis Etherchannel"
The specced kit for the core are 3850's, in an ideal world I'd use VSS (Virtual Switch System) to achieve the above statements beyond repute; but this is only supported on 4500/6500 and Nexus platforms.
Do we think a cross stack etherchannel (LACP between both core switch stacks) would satisfy the above statements? Or the statements may just be badly worded...
I look forward to your thoughts and views on this! Thanks!

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
As the others have noted, the 3850s, to stack, are restricted to the length of the longest stack cables.
As you have noted, VSS physical units would allow the "logical" unit to be far apart.
For a "small" VSS core, the 4500-X might be an idea unit. (Other than cost, the 4500 would be a better choice for a core device.)
Something to watch for, or understand, when running VSS, Etherchannel doesn't load balance as it does on a single chassis or stack. VSS will avoid using the VSL cross link unless it must.
As many access switches, today, support basic L3 routing, you might also determine whether a L3 edge would be a suitable alternative choice. It would allow retention of the 3850s and can offer some advantages even over VSS. (Where VSS is very nice [as too the Nexus] supporting servers with Etherchannels.)

Checking the device ASM picked up when multi-pathed.

Hi,
I may have missed this in the documentation but I'd can't find out how to comfirm that ASMlib has picked up the multi-path device rather than one of the real paths where storage is multi-pathed.
Is there a way to confirm the device found by ASMlib?
The environment is 11g on Linux using ASMlib.
The search string should get ASM to find the mulit-path device first but I would like to ocnfirm what has actually happened.
Many thanks in advance.

Hi,
I've been scanning metalink again for inspiration on this and found the document "FAQ ASMLIB CONFIGURE,VERIFY, TROUBLESHOOT" (Doc ID: 359266.1).
I don’t know why I didn’t find it earlier but that’s life.
By the way that is a useful doc to read.
This solution answers my question of “How do I know what device ASMlib has actually found and that it is using the multipath device?”.
First find the major,minor numbers for the device ASMlib is using.
# /etc/init.d/oracleasm querydisk -d VOL4
Disk "VOL4" is a valid ASM disk on device [251, 97]
Next check the major,minor numbers against the device in /dev.
# ls -l /dev/* | grep 251 | grep 97
brw-r----- 1 root disk 251, 97 Jun 11 19:12 /dev/sddlmag1
Simple, I just hadn’t got it until now.
Thanks for the suggestions posted.
Edited by: user10515008 on 12-Jun-2009 04:08
Removed the bolding on the examples as it confused the actual display.

Recmd for low cost, multi point temp logging

Can someone recommend the best, low cost way to monitor temp at 16
pts. We are doing qualifiation tests of an oven used in an bio tech
research application. But we are a startup and I need to stretch my
capital budget as much as possible. Here are the details:
1. type k theromocouople
2. 40-150F temp range
3. 16 tc input channels
4. scan rate = 1 reading / minute
5. log period = 60 hrs
I used National Instruments Field Point modules about 2 yrs ago for
this since they were the cheapest approach I found for the signal
condtioning since a 8 channels of type K tc inputs, the TC120 module,
cost $350.
Is there a better approach I should look at? I do not have to work
with LV or NI hw if there is a better approach.
Thanks in adv
ance for your advice,
Doug Danielson
[email protected]
Atlanta, GA
recmd for low cost, multi point temp logging

On 18 Jul 2003 14:59:07 -0700, [email protected] (William
Danielson) wrote:
>Can someone recommend the best, low cost way to monitor temp at 16
>pts. We are doing qualifiation tests of an oven used in an bio tech
>research application. But we are a startup and I need to stretch my
>capital budget as much as possible. Here are the details:
>
>1. type k theromocouople
>2. 40-150F temp range
>3. 16 tc input channels
>4. scan rate = 1 reading / minute
>5. log period = 60 hrs
>
>I used National Instruments Field Point modules about 2 yrs ago for
>this since they were the cheapest approach I found for the signal
>condtioning since a 8 channels of type K tc inputs, the TC120 module,
>cost $350.
>
>Is there a better approach I should look at? I do not have to
work
>with LV or NI hw if there is a better approach.
>
>Thanks in advance for your advice,
>
>Doug Danielson
>
>[email protected]
>Atlanta, GA
>
>
>recmd for low cost, multi point temp logging
I don't remember where I got them from but I had a similar project a
few years ago and found that thermistors are a lot cheaper (and
slightly less accurate) than thermocouples. get an AtoD card for a pc
and you are pretty much in buisness. You'll need some resistors too
for voltage dividers. You need to watch what voltage you put across
them so you don't get significant internal heating. The suppliers
have the schematics and other usefull info. Thermistors don't give
you a linear reaction, so you need to use a table, graph, or an
emperical equation get actual temps. You will also want to calibrate
all of them as there will be some variance between sensors (even if
they were perfect the resistors you use along with them will vary)
rather than try to get them all dead nuts on with li
ttle trimpots just
figure a fudge factor for each sensor for when you calculate back the
actual temps. I'd try them all in ice water and boiling water, just
seal them in plastidip or something.before immersing them.
David

If equal cost routes exist, OSPF uses CEF load balancing?

Hi All,
Can anyone explain about:
. If equal cost routes exist, OSPF uses CEF load balancing?

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
Rick is correct, but if his response, with mine, causes any confusion. . .
To OP's original question:
If equal cost routes exist, OSPF uses CEF load balancing?
The answer is technically no, for the reason Rick describes.
But if we rephrase, such as:
Does CEF load balance across multiple equal cost routes generated by OSPF?
The answer would be yes.
I suspect the latter question is what the OP really had in mind, but again, Rick is correct to distinguish that OSPF doesn't use CEF.

OSPF Equal Cost (multiple links) but unbalanced traffic

Hi!
I would like to ask about OSPF.
We are using 4 links running OSPF point to point with equal cost to load balance MPLS L2 traffic between our two routers.
And from what we are experiencing right now is that it didnt balance the load the way it should have.
The first link always catches up the bulk of the traffic, while the remaining 3 links load balances, I would like to inquire on steps to mitigate the issue.
Appreciate if I could receive some help to solve the issue
Thank you very much in advance!
Best Regards,
asakurahao

I do have almost the same problem.
i have 2 x 20 Mbps Point to Point ethernet links configured with equal ospf cost and both links terminated on the single router at both ends.
interface FastEthernet0/0/0
bandwidth 20000
ip address x.x.x.x y.y.y.y
load-interval 30
duplex auto
speed auto
traffic-shape rate 20000000 500000 500000 1000
interface FastEthernet0/0/1
bandwidth 20000
ip address x.x.x.x y.y.y.y
load-interval 30
duplex auto
speed auto
traffic-shape rate 20000000 500000 500000 1000
FastEthernet0/0/0 is up, line protocol is up
Internet Address 135.254.193.53/30, Area 0.0.0.1
Process ID 3435, Router ID 135.254.0.95, Network Type BROADCAST, Cost: 5
FastEthernet0/0/1 is up, line protocol is up
Internet Address 135.254.193.45/30, Area 0.0.0.1
Process ID 3435, Router ID 135.254.0.95, Network Type BROADCAST, Cost: 5
I see majority of the traffic using the link F0/0/0 and only 30% traffic using the second link F0/0/1.
I used per-packet load-sharing on these interfaces which caused throughput problems, so i removed that from the interface. Should i add some CEF commands or should try per-destination load sharing to see equal traffic on both links ?

OSPF Equal Cost Path Selection

This is a nerdy enough qeury in reality.
We have a single area - area 0.0.0.32. All intra-area routes. We have 2 switches in the core of the network, and 10 switches at the edge. All of these switches are connected via layer 3 OSPF routed links.
The cost for all links is 20 - which is based on bandwidth between the boxes - which is 2Gbps.
Have a look at the enclosed jpeg to get an idea.
Very simple.
Query revolves around the path selection available to OSPF.
The path from Core 2 to the 10.32.51.0 network is easy - straight across the link between the core switches for a cost of 20.
Question :- if the link between the cores fail, which path will be chosen by OSPF and why?
It can go through ANY of the other edge switches for a total cost of 40, but it will choose a particular one. What criteria does OSPF use to select this path?
Remember, the path costs are equal, they are all intra-area.
I tried messing with Router ID, but this doesn't seem to be it.
I tried highest interface IP addresses, but it doesn't seem to be this.
It is not a random act, the algorithm chooses the same one every time.
There must be some parameter in the LSDB that is the defining one when it comes to path selection.
Can you help me out please. I need someone who understands the OSPF algorithm better than I do (which might not be hard!).
Appreciate any comments for debate.

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
My guess (as I haven't re-read the RFC), selection of an ECMP to retain in a routing table (assuming all possible ECMP are not retained) and/or exactly how packets or flows are ECMP routed is implementation dependent.
I recall years ago bumping into a situation where I had 6 (OSPF) ECMP on a Cisco router which had the (then) default allowance of 4 ECMP in the route table. I don't recall exactly what the issue was, but whatever it was doing I considered it a bug. The "fix" was to allow the router to use all 6 ECMP. (Again, don't remember the specifics, but the issue I saw was more involved than 2 of the 6 ECMPs weren't retained.)

MPLS and dual equal-cost paths

If a site has two outbound WAN routers (for redundancy) that are configured as PE routers with the WAN interfaces configured with MPLS, will two LSPs be created using LDP (one from each PE) allowing the WAN routers to loadbalance traffic across the MPLS backbone?

Load balancing is very much possible in MPLS environment. Following are the general rules for load balancing labeled packets
1. If the MPLS payload is an IPV4 packet, the load balancing is done by hashing the source and destination IP address of the IPv4 header
2.IF the MPLS payload is an IPv6 packet, the load balancing is done by hashing the source and destination IP address of the IPv6 header.
3.If the MPLS payload is not an IPv4 packet, the load balancing is done by looking at the value of the bottom label
Label stack does not have protocol identifier field so to know what is the MPLS payload is, router can look first nibble following the MPLS label stack, if it is 4 then router considers this an IPv4 packet and performs IPv4 CEF hashing. if it is 6 , MPLS payload is considered IPv6, and the router performs IPv6 CEF hashing
For more information you can refer book "MPLS fundamentals" by Luc De Ghein.

Actual Costing - Multi Level

Hi,
Request the team to revert , to perform the variance (Price & Usage) allocation to the multi level cost roll up from the inter-mediate stage to the Finished Goods stage, need to know the method to perform the same.
Thanx
Ganesh

Hi,
The variances can be transferred to COPA /PCA.The variances should not be carried forward to the finished goods.[ you should not manually post the variances to the finished goods a/c].
If you use COPA,you should use a transfer structure [ PA structure] to settle the variances-this structure can be maintained in order type etc.in the M/data.
If you assign all the org-units to PCA,the variances can be got from the reports.
Hope this helps.
Regards,
Ramesh

ECMP - equal cost multi path

Similar Messages

Maybe you are looking for