Go Premium for a chance to win a PS4. Enter to Win

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 303
  • Last Modified:

bgp an inbound traffic

I have a multihoming setup with two different ISPs. It is a primary (ISP A) and backup (ISP B) type of setup. I am using AS prepending on ISP B. Now the problem that I am having is the inbound traffic. I am getting all of the inbound traffic via my ISP B. Is there a way to have some goes to ISP A? Thanks
0
leblanc
Asked:
leblanc
  • 4
  • 3
  • 2
  • +2
8 Solutions
 
Garry GlendownConsulting and Network/Security SpecialistCommented:
Can you give some more details? What steps have you taken to figure out what is causing traffic to come in via IPS B? How many Prepends did you add to the announcements to ISP B? Prepends do not guarantee that there won't still be traffic reaching you via the uplink, but just tip the scales so to speak ...
0
 
btanExec ConsultantCommented:
thought this is a good post of basic - eventually it is the way the route is broadcast and BGP only advertises the best path for any given prefix to its neighbors. Hence you tend to see only only path even though that may not be the ideal.....then again the author state something on having a load balance traffic which I extracted below for your info...

http://www.networkengineerdigest.com/2010/12/22/bgp-as-path-prepending/

Also, a note about load balancing.  There is no possible way to create a policy that load balances BGP traffic in a 50/50 manner.  There are path selection rules that always come down to preferring a one path over another.  The more source and destinations in BGP the better the distribution of traffic.  If you have a single route to advertise – then all return traffic will mostly take a single link (aka the best path).  If you receive a single default route from your upstream peer, then all your traffic will take a single link from your network.  More advertised routes from your network the better the return traffic to your network – More routes received from your upstream peers the better the outbound distribution of traffic from your network.
0
 
pergrCommented:
Which size networks do you have?

In case you have two /24, then you can advertise two /24 via ISP A and a /23 via ISP B. That way all traffic should come via A regardless of prepends.

Perhaps now A is not even advertising your network to the world.
0
Free Backup Tool for VMware and Hyper-V

Restore full virtual machine or individual guest files from 19 common file systems directly from the backup file. Schedule VM backups with PowerShell scripts. Set desired time, lean back and let the script to notify you via email upon completion.  

 
giltjrCommented:
If you have a single /24, not much you can do.

If you have two /24's, then what pergr suggested will work as long as you have have hosts with IP addresses within both ranges.

If you have a single /24, which ISP gave it too you?

We do BGP with our ISP's and advertise a single /24.  Both our ISP's update their routing tables so that any host that is part of their network uses their link to get to us.  So even though we may prefer ISP#1 based on our BGP parameters, if somebody is directly connected to ISP#2, they use ISP#2's link.
0
 
leblancAccountingAuthor Commented:
I found out from ISP B that I am overutilizing the connection. I prepended the BGP as_path 5 times on ISP B.

I do not do load balance. I have a fairly small network with 10 remote sites.

ISP A is my primary MPLS and Internet connection. ISP B is my backup internet. But on the inbound, it looks like everything is going to ISP B.

I have only one /24 public IP address from ARIN.

Thanks
0
 
pergrCommented:
It seems like that ISP A is doing something wrong, and are not advertising your network to their uplinks.

You should look in some "Looking Glass" sites, what AS PATH you can see to your /24.
0
 
btanExec ConsultantCommented:
This spells out issue on "lots of" AS prepends but this wouldnt necessarily be the cuplrit unless there is limits configured. Most of the time is seems alright
http://blog.packetsource.net/2009/02/21/problems-with-bgp-prepending/

Excerpt: When configuring the prepend operation, instead of specifying the desired AS path to be seen – as might have been expected on a Cisco IOS device – the configuration asked the operator to specify the number of times to prepend instead.

Coming back, normally, with as-path prepending on outgoing Ebgp updates through the non-desired return path and having as-paths sent out over the non-desired link, it will (or should) become longer than the as-path directed to the preferred path. However, we also must remember that BGP route selection uses couple of selection criteria (not limited to the list below) in descending order of preference.
i.e. LocalPreference has priority over on as-path length.

i. Prefer the higher-value weight.
ii. Prefer the higher-value local preference.
iii. Prefer routes that the router originated.
iv. Prefer shorter length of AS paths.

Below has a link that resolved via a community value (for local pref) over as path (as prepending). Not sure if this can help
http://networktalk.wordpress.com/bgp-tutorial-multihoming-practices-cisco/chapter-b-study-cases-traffic-engineering-concerns-setup-and-implementation-of-ebgp-and-ibgp-sessions/b-1-traffic-engineering-tools-discussions-for-multihoming-purposes/b-1-a-choice-of-bgp-tools-and-practices-for-return-traffic-as-path-prepend-localpreference-or-bgp-communities/
0
 
btanExec ConsultantCommented:
Useful info from CISCO

Load Sharing When Multihomed to Two ISPs Through a Single Local Router
http://www.cisco.com/en/US/tech/tk365/technologies_configuration_example09186a00800945bf.shtml#conf4
0
 
Garry GlendownConsulting and Network/Security SpecialistCommented:
Did you check on some LG servers or other sources how your prefixes look from the outside? What happens if ISP B goes down - are you still able to get everywhere? If so, then it may be something with the prepends. If not, then ISP A (or their Uplink(s)) is most likely filtering your prefix or your AS ... could happen when the uplinks do filtering by DB from places like RIPE etc, and ISP A didn't update their data with your AS ...
0
 
pergrCommented:
On option is, of course, that ISP A is actually a customer of ISP B..., in which case, as the traffic arrives to ISP B, they will prefer the direct route to you (based on local preference) instead of sending it to ISP A.

Instead of doing prepends, you can check if your ISPs are offering some sort of bgp community based engineering. Effectively, you would send them specific community values, and based on those the ISP will prepend to their uplinks - instead of you adding the prepends  yourself.
0
 
pergrCommented:
As a test, you can shut down the ISP B link temporarily, and see if the traffic comes to ISP A.
0
 
leblancAccountingAuthor Commented:
I will test that
0

Featured Post

Free Tool: Subnet Calculator

The subnet calculator helps you design networks by taking an IP address and network mask and returning information such as network, broadcast address, and host range.

One of a set of tools we're offering as a way of saying thank you for being a part of the community.

  • 4
  • 3
  • 2
  • +2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now