In Feburary of this year VMware released NSX-T 2.0 and with it came a variety of updates that looked to continue to push of NSX-T beyond that of NSX-v while catching up in some areas where the NSX-v was ahead. The NSBU has big plans for NSX beyond vSphere and during the NSX vExpert session we saw how the future of networking is all in software…having just come back from AWS re:Invent I tend to agree with this statement as organisations look to extend networks beyond traditional on-premises or cloud locations.
NSX-T’s main drivers relate to new data centre and cloud architectures with more hetrogeneality driving a different set of requirements to that of vSphere that focuses around multi-domain environments leading to a multi-hypervisor NSX platform. NSX-T is highly extensible and will address more endpoint heterogeneity in future releases including containers, public clouds and other hypervisors. As you can see before the existing use cases for NSX-T are mainly focused around devops, micro-segmentation and multi-tenant infrastructure.
Layer 3 accessibility across all types of platforms.
What’s new in NSX-T 2.1:
Today at Pivotal SpringOne, VMware is launching version 2.1 of NSX-T and with it comes a networking stack underpinning Pivotal Container Services, direct integration with Pivotal Cloud Foundry and significant enhancements to load balancing capabilities for OpenStack Neutron and Kubernetes ingress. These load balancers can be virtual or bare metal. There is also native networking and security for containers and Pivotal operations manager integration.
NSX-T Native Load Balancer:
NSX-T has two levels of routers as shown above…then ones that connect to the physical world and the ones which are labeled T1 in the diagram above. Load balancing will be active on the T1 routers and have the following features:
- Algorithms – Round Robin, Weighted Round Robin, Least Connections and Source IP Hash
- Protocols – TCP, UDP, HTTP, HTTPS with passthrough, SSL Offload and End to end SSL
- Health Checks – ICMP, TCP, UDP, HTTP, HTTPS
- Persistance – Source IP, Cookie
- Translation – SNAT, SNAT Automap and No SNAT
As well as the above it will have L7 manipulation as will as OpenStack and Kubernetes ingress. Like NSX-v these edges can be deployed in various sizes depending on the workload.
Pivotal Cloud Foundry and NSX-T:
For those that may not know, PCF is a cloud native platform for deploying and operating modern applications and in that NSX-T providers the networking to support those modern application. This is achieved via the Network Container Plugin. Cloud Foundry NSX-T topology include a separate network topology per orginization with every organization getting one T1 router. Logical switches are then attached per space. High performance north/south routing uses NSX routing infrastructure, including dynamic routing to the physical network.
For east/west traffic that happens container to container with every container having distributed firewall rules applied on it’s interface. There is also a number of visibility and troubleshooting counters attached to every container. NSX also controls the IP management by supplying subnets from IP blocks to namespaces and individual IPs and MACs to containers.
Log Insight Content Pack:
As part of this release there is also a new Log Insight NSX-T Content Pack that builds on the new visibility and troubleshooting enhancements mentioned above and allows Log Insight to monitor a lot of the container infrastructure with NSX.
When it comes to the NSX-T 2.1 feature capabilities, the load balancing is a case of bringing NSX-T up to speed to where NSX-v is, however the thing to think about is that how those capabilities will or could be used beyond vSphere environments…that is the big picture to consider here around the future of NSX and it can be seen with the deeper integration into Pivotal Cloud Foundry.