VMworld: Networking Deep Dive
At VMworld I attended the excellent “Networking Deep Dive” (TA2525) by Srinivas Neginhal. Here are just a few things he discussed that jumped out me as either new information (not in the standard documentation) or something that expanded on my current understanding.
ESX hosts keep a local cache of the vDS and DVPort information, to use when vCenter is unavailable (both are non-editable binary files):
Host DVPort state: /etc/vmware/dvsdata.db
VM DVPort state: /vmfs/volumes/<storage>/.dvsData
DVPort binding types:
- Static – assigned to the virtual adapter and immediately pushed to the host. Written to the host’s cache and written in the VM’s vmx file.
- Dynamic – assigned when the VM is powered on, and then pushed to the host. There is no guarantee that the VM will get the same DVPort on the next power on. However it uses a concept similar to DHCP in that if the same port is available then it will renew that one.
- Ephemeral – a new port on every power-on. The port is destroyed when the VM disconnects from the port. Essentially this is what happened on all vSwitch ports on ESX 3.x and lower. This is the only option if you are connecting the client directly to the host, vCenter is required to create the other 2 types of binding.
He also explained the Traffic Shaping terminology, which seems obvious in retrospect but was something that I hadn’t “got”. Simply put, its all relative to the switch. So Ingress traffic is stuff coming in, whether from a VMNIC (physical adapter) or a vNIC (VM’s adapter). Egress traffic is leaving the switch.
And one last thing which stood out was a table he used to quickly show some of the advanced vNIC driver capabilities. I’d not seen this laid out in this way, and is much more specific than the standard documentation (in fact if this is correct, which I’m sure it is, then the current documentation is downright misleading – N.B I need to update my vSphere reference card with this). It also makes it much easier to remember.
TSO – TCP Segmentation Offloading
RSS – Receive Side Scaling http://www.microsoft.com/whdc/device/network/NDIS_RSS.mspx
MSI/MSI-X – Message Signaled Interrupts http://blogs.sun.com/gnunu/entry/how_does_msi_x_work
Update: VMware has just released this performance whitepaper discussing the merits of the new vmxnet3 driver.
2 Responses to VMworld: Networking Deep Dive
Leave a Reply Cancel reply
Forbes Guthrie
Recent Posts
- Small update to the Reference Card
- Minor update to the vSphere 5 Reference Card
- vSphere 5 vReference Card released
- Cisco UCS boot from iSCSI SAN – ESXi design consideration
- vSphere 5 vReference card – Storage section
- Does 2008 R2 Failover Clustering require a change to the Notify Switches policy?
- vSphere 5 vReference card – Host section
- vSphere 5 vReference card – Install section
- Auto Deploy design concern
- vSphere 5 vReference card – vCenter section
Recent Comments
- honglus on How to PXE boot from your trunked vmnic0
- Sunmeet on Understanding ESXi – stateless, diskless, feckless
- Forbes Guthrie on vSphere 5 Card
- Forbes Guthrie on vSphere 5 Card
- Jakk on vSphere 5 Card
- Purushothama S on vSphere 5 notes
- Bjorn on vSphere 5 Card
- Chris on Minor update to the vSphere 5 Reference Card
- Michael Webster on Auto Deploy design concern
- Ankur Maheshwari on vSphere 5 notes
Twitter
- As much as I've become an NFS supporter over the years, VMFS5 & ATS is really making me like iSCSI & its better multipathing. : 2 weeks ago
- @__wintertale__ Hi Iona (& Lesley). Welcome to the twitterverse from a cloudy and wet Vancouver. How's the coffee? : 3 weeks ago
- RT @jtroyer: ... Kaua'i http://t.co/s4Ovo4jD << Staying in same resort next month. Pls, no destroying it with wild parties ;) : 3 weeks ago
- RT @dobaer: @forbesguthrie How was the run in your shiny new shoes? << Nice, the trails around Vancouver are stunning. : 3 weeks ago
- @h0bbel I'm a Salomon addict, but I'm scared to get these ones muddy :) : 3 weeks ago
- Just took my pwutty new #salomon trail shoes for an hour's run in Pacific Spirit Park. http://t.co/7dAQDXgl : 3 weeks ago
- RT @alim__k: @forbesguthrie @keith_aasen look at you guys, figuring this out over social media ;-) < that's where I do all my best designs : 3 weeks ago
- @keith_aasen Great, just what I was after. And presumably one per controller to prevent performance degradation on failover (after refill) : 3 weeks ago
- @keith_aasen You're the VDI sizing TR author :) You're my Guru! : 3 weeks ago
- NetApp gurus: VDI on 3170 - where's FlashCache watershed for 2x 512GB to 4. When is it 2TB not 1TB? Don't say "it depends" /cc @keith_aasen : 3 weeks ago







Hi Forbes,
Just found your site while looking for information on ports used for communications between vCenter and ESX – really good stuff, thanks for sharing!
In this article the table has IPv4 TSO twice – am I correct in assuming that the second line should be IPv6 TSO?
Thanks again.
Hi Chris, yes you’re quiet right, the second line should be IPv6 TSO. I’ll update it soon as I can. Thanks for the heads up.