<div dir="ltr">Wondering if you might have some imbalance in a LAG somewhere. Where it is hashing too much traffic to one link of a lag. By default it uses the mac address of the next layer 2 hop and traffic going to a gateway will all hash to the same link. Are there any LAGs involved? Is there a major imbalance of traffic on a LAG in the traffic path?</div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Feb 12, 2015 at 9:43 PM, <span dir="ltr"><<a href="mailto:nethub@gmail.com" target="_blank">nethub@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">We are only accepting about 300k IPv4 routes currently (we filter to reduce<br>
the table size). We are on the multi-service-2 CAM partition profile and we<br>
have the system-max values for ip-route and ip-cache set to 445K.<br>
<br>
Also, we upgraded to 5.6f today to see if that would help but it did not<br>
change anything.<br>
<br>
CPU usage is very low across the board (under 10% use on everything), so if<br>
it is routing in software, it isn't causing a jump in CPU load.<br>
<span class="im HOEnZb"><br>
<br>
-----Original Message-----<br>
From: foundry-nsp [mailto:<a href="mailto:foundry-nsp-bounces@puck.nether.net">foundry-nsp-bounces@puck.nether.net</a>] On Behalf Of<br>
</span><div class="HOEnZb"><div class="h5">Niels Bakker<br>
Sent: Thursday, February 12, 2015 8:38 PM<br>
To: <a href="mailto:foundry-nsp@puck.nether.net">foundry-nsp@puck.nether.net</a><br>
Subject: Re: [f-nsp] MLX throughput issues<br>
<br>
* <a href="mailto:nethub@gmail.com">nethub@gmail.com</a> (<a href="mailto:nethub@gmail.com">nethub@gmail.com</a>) [Fri 13 Feb 2015, 01:45 CET]:<br>
>As I stated in the first message, the Juniper EX3200 is a downstream<br>
>BGP customer that is single homed to our network, so it is on a<br>
>different ASN and the communication between my network and his network<br>
>is layer 3.<br>
<br>
Are you running that MLX with a full BGP table? 20 MB/sec sounds like<br>
you're forwarding packets over its CPU, perhaps because it ran out of CAM<br>
space.<br>
<br>
<br>
-- Niels.<br>
<br>
--<br>
_______________________________________________<br>
foundry-nsp mailing list<br>
<a href="mailto:foundry-nsp@puck.nether.net">foundry-nsp@puck.nether.net</a><br>
<a href="http://puck.nether.net/mailman/listinfo/foundry-nsp" target="_blank">http://puck.nether.net/mailman/listinfo/foundry-nsp</a><br>
<br>
_______________________________________________<br>
foundry-nsp mailing list<br>
<a href="mailto:foundry-nsp@puck.nether.net">foundry-nsp@puck.nether.net</a><br>
<a href="http://puck.nether.net/mailman/listinfo/foundry-nsp" target="_blank">http://puck.nether.net/mailman/listinfo/foundry-nsp</a><br>
</div></div></blockquote></div><br></div>