[rbak-nsp] SE400 link aggregation problem with load balancing

[Elma K.]

Hi :)

I our situation this should work like a charm, because those are transit links from our subscribers (some 10K of them) to Internet...
Plenty of IPs :)... But it's not working. For a temporary solution I configured 4 link groups with 2 ports in each, so I have 800M
download (instead of 600M before), and 400M available for upload (because upstream traffic is going through only one port in group).
This would hold for some time until I find some other solution :(

Regards.
Elma



Subject: Re: [rbak-nsp] SE400 link aggregation problem with load balancing
From: ripleydotnet at gmail.com
Date: Thu, 21 Jan 2010 13:47:24 -0700
CC: tanan.satayapiwat at ericsson.com; frans at falco-networks.com; redback-nsp at puck.nether.net
To: elmicak at hotmail.com



Yes, old SEOS version.  LACP was introduced much later than 2.6.7
However, the hash should be source/dest IP, so a single (or a few) source/destination pair's will always pick one link, it will take 100's of pairs for the hashing algorithm to work for load balancing.
HTH

On 2010-01-20, at 11:37 PM, Elma K. wrote:Unfortunately, there are no commands for lacp protocol under link group... Possible because old SEOS version.

Subject: Re: [rbak-nsp] SE400 link aggregation problem with load balancing
From: ripleydotnet at gmail.com
Date: Wed, 20 Jan 2010 12:12:44 -0700
CC: tanan.satayapiwat at ericsson.com; frans at falco-networks.com; redback-nsp at puck.nether.net
To: elmicak at hotmail.com

make sure you enable lacp active in the link-group.
On 2010-01-20, at 8:53 AM, Elma K. wrote:Source IPs on SE400 side are from the pool configured on that device, and destination IPs are various (Internet traffic for our pppoe subscribers)...
Link IPs should not have meaning in this case, because link is just for transit traffic. Also, source and destination IPs should be different, because
it is the most unlikely that this traffic is from just one subscriber sending to just one IP...


From: tanan.satayapiwat at ericsson.com
To: elmicak at hotmail.com; frans at falco-networks.com; redback-nsp at puck.nether.net
Date: Wed, 20 Jan 2010 10:13:52 -0500
Subject: RE: [rbak-nsp] SE400 link aggregation problem with load balancing

Frans mean the source and destination IP of the packets traveling through the link. It does not have to be IP address of the link itself since the traffic can be transit traffic. Tanan
From: redback-nsp-bounces at puck.nether.net [mailto:redback-nsp-bounces at puck.nether.net] On Behalf Of Elma K.
Sent: Wednesday, January 20, 2010 4:10 PM
To: frans at falco-networks.com; redback-nsp at puck.nether.net
Subject: Re: [rbak-nsp] SE400 link aggregation problem with load balancing

The link is with /30 mask, so I can not take just 1 higher IP... Maybe I can change both IPs on this link, 
but I have to check this with a coleague that works on the other device, to see if this is possible for him.

I will let You know when I get this info. In the meantime, if there is any other idea, I would appreciate it.

Regards
Elma



Date: Wed, 20 Jan 2010 16:03:25 +0100
Subject: Re: [rbak-nsp] SE400 link aggregation problem with load balancing
From: frans at falco-networks.com
To: elmicak at hotmail.com; redback-nsp at puck.nether.net

Hi Elma,

Do you by any change use the same IP address as destination and/or source?
The way the balancing works is based on something like source IP plus destination IP plus destination port, that will stick to the same physical link.
I ran into this problem once, solved the test by adding another source 1 IP address higher and that took another link. 

I don’t know if you have the same kind of trouble there.

Kind regards,


Frans.


From: "Elma K." <elmicak at hotmail.com>
Date: Wed, 20 Jan 2010 14:57:15 +0000
To: <redback-nsp at puck.nether.net>
Subject: [rbak-nsp] SE400 link aggregation problem with load balancing

Hello.

I have SE400 connected to other device, and I have configured link aggregation on SE400, 
everything works fine except load-balancing from Redback's side... 
Traffic that is going from redback is not balanced on link group members, it is gong only on one port.

I guess there must be some configuration to change the way the device is doing this load balancing
between group members.

My device is with SEOS-2.6.7 (no chance for upgrade, unfortunately)
and the config is like this:

link-group hsiUplink-2PCh ether
 bind interface uplink-2 hsi

port ethernet 1/5
 link-group hsiUplink-2PCh
!
port ethernet 1/7
 link-group hsiUplink-2Ch
!
port ethernet 3/12
 link-group hsiUplink-2PCh
!

Could anyone help me, please?

Best Regards.
Elma
       
Windows Live:  Friends get your Flickr, Yelp, and Digg updates when they e-mail you. <http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/social-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_3:092010> 
_______________________________________________
redback-nsp mailing list
redback-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/redback-nsp

Keep your friends updated— even when you’re not signed in. 
Windows Live: Friends get your Flickr, Yelp, and Digg updates when they e-mail you._______________________________________________
redback-nsp mailing list
redback-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/redback-nsp



--Ron Ripley, Network Architectripleydotnet at gmail.com




Windows Live: Make it easier for your friends to see what you’re up to on Facebook.



--Ron Ripley, Network Architectripleydotnet at gmail.com





 		 	   		  
_________________________________________________________________
Windows Live: Keep your friends up to date with what you do online.
http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/social-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_1:092010
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/redback-nsp/attachments/20100122/37b7cab4/attachment-0001.html>