<div dir="ltr">Hey guys! Any ideas about how to load balance or split the VXLAN network traffic?</div><div class="gmail_extra"><br><div class="gmail_quote">On 10 April 2015 at 01:39, Martinx - ジェームズ <span dir="ltr"><<a href="mailto:thiagocmartinsc@gmail.com" target="_blank">thiagocmartinsc@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hello Remo, Erik,<div><br></div><div> Thanks for this fast reply! :-)</div><div><br></div><div> Remo, I want to seg the communication between my Instances (i.e., intra-vm communication), for example, I want to "scp from-instance-a to-instance-b" through "net-a" and simultaneously, "scp from instance-b to-instance-a" through "net-b", so, there will be no bottleneck. On this example, both Instances have two subnets attached but, I would like to use two VXLAN tunnels, one for "net-a" and another for "net-b"... And I'm not sure if this is even possible to achieve... Is it?</div><div><br></div><div> Evolving this scenario, I'll build something like this:</div><div><br></div><div> * Instance-A -> net-a -> "Instance-Bridge" -> net-b -> "Instance-Gateway" -> Internet</div><div><br></div><div> Here, the "Instance-bridge" will be connected at both "net-a" and "net-b", and will act as bridge. The gateway IP of "Instance-a" is allocated at "Instance-Gateway".</div><div><br></div><div> So, I'll need to, somehow, attach "net-a", to a VXLAN tunnel associated with "segment 1", and "net-b", to a VXLAN tunnel associated with "segment 2".</div><div><br></div><div> Erik, in fact, no, I don't need more bandwidth, I just need to make sure that the network traffic going in and out from "Instance-bridge / Instance-Gateway" will face no bottleneck. I'm thinking about the possibility to build a LACP BOND channel, using 4 ethernets, for just one VXLAN subnet. This way, I believe it will more dynamic and easy to build/manage... What do you think? Ah, and yes, I'm intending to segment the traffic in a specific way, as my above example.</div><div><br></div><div> Thanks again!</div><div><br></div><div>Best,</div><div>Thiago</div></div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><div class="gmail_quote">On 9 April 2015 at 23:32, Erik McCormick <span dir="ltr"><<a href="mailto:emccormick@cirrusseven.com" target="_blank">emccormick@cirrusseven.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><p dir="ltr">If your goal is to simply have more available bandwidth, I would think you could bond the 4 interfaces into one and enable jumbo frames. Are you intending to segment the traffic for some other purpose?</p>
<div class="gmail_quote"><div><div>On Apr 9, 2015 7:20 PM, "Martinx - ジェームズ" <<a href="mailto:thiagocmartinsc@gmail.com" target="_blank">thiagocmartinsc@gmail.com</a>> wrote:<br type="attribution"></div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div dir="ltr">Hey guys!<div><br></div><div> I'm looking to build an OpenStack environment, where I have 4 physical interfaces that will be dedicated to the VXLAN tunnels but, how can I do that?</div><div><br></div><div> For example, I planning to create 4 "net / subnets" and each one of it will be associated with 1 VXLAN tunnel (of 4?)...</div><div><br></div><div> My need is that some traffic will come from one segment, pass through a Instance and goes out through another segment. So, in fact, I need to balance the network traffic, throughout more than one segment, to avoid bottlenecks.</div><div><br></div><div><br></div><div> Which approach better fits into my needs?</div><div><br></div><div> A- 1 VXLAN tunnel for each ethernet segment (4 tunnels) (is this even possible?)? or</div><div><br></div><div> B- 1 VXLAN on top of a BOND channel (here, the BOND channel will be built using 4 ethernet segments)?</div><div><br></div><div><br></div><div> I appreciate any tip! :-)</div><div><br></div><div> Thanks in advance!</div><div><br></div><div>Cheers!</div><div>Thiago</div></div>
<br></div></div><span>_______________________________________________<br>
Mailing list: <a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack</a><br>
Post to : <a href="mailto:openstack@lists.openstack.org" target="_blank">openstack@lists.openstack.org</a><br>
Unsubscribe : <a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack</a><br>
<br></span></blockquote></div>
</blockquote></div><br></div>
</div></div></blockquote></div><br></div>