Nicira, Inc.

Vmware, Inc.

Nilesh Ramchandra Nipane

Jia Yu

Sachin Thakkar

Piyush Kothari

ChiHsiang Su

Nikhil Ravindra Rajguru

Rushikesh Shashank Ghatpande

Pankaj Thakkar

Ganesan Chandrashekhar

Abhishek Goliya

Akshay Katrekar

Chi-Hsiang Su

Lele Zhang

Maheedhar Nallapareddy

Kiran Desai

Shyam Sambasivan Ramachandran

Aarti Lolage

Akshay Kale

Kaustubh Naniwadekar

Janani Natarajan

Lele Zhang Zlele

Jobin John

Example methods and systems for priority-based network bandwidth allocation are described. In one example, a first computer system may detect an event indicating that network bandwidth allocation is required for a virtualized computing instance. The first computer system may identify, from multiple priority levels, a first priority level that is associated with (a) the virtualized computing instance, (b) a logical network element to which the virtualized computing instance is attached, or (c) a group that includes the virtualized computing instance or the logical network element. The first computer system may obtain network bandwidth capacity information associated with physical network adapter(s) capable of forwarding traffic associated with the virtualized computing instance. Based on the first priority level and the network bandwidth capacity information, the first computer system may configure a priority-based network bandwidth allocation policy that includes parameter(s) applicable to the traffic associated with the virtualized computing instance.

Priority-based network bandwidth allocation

The disclosure provides an approach for virtual computing instance (VCI) migration. Embodiments include scanning logical segments associated with a customer gateway to identify network addresses associated with the logical segments. Embodiments include determining one or more recommended supernets based on the network addresses associated with the logical segments. Embodiments include providing output to a user based on the one or more recommended supernets. Embodiments include based on the output, receiving input from the user configuring an aggregation supernet for the customer gateway. Embodiments include advertising the aggregation supernet to one or more endpoints separate from the customer gateway.

Supporting virtual machine migration when network manager or central controller is unavailable

The disclosure provides an approach for in-place conversion of a virtual switch on a host. Techniques are provided for in-place conversion of a virtual switch from a first type of virtual switch to a destination type of virtual switch. A method includes rekeying, by a second manager agent, one or more uplink ports associated with one or more logical switches implemented by the virtual on the host with the unique identifier. The rekeying includes updating one or more existing uplink port identifiers assigned to the one or more uplink ports with the unique identifier associated with a first manager agent. The method includes removing, by the first manager agent, an association of the one or more uplink ports with the opaque network; and informing the first manager and the second manager that the virtual switch is of the destination type of virtual switch.

In-place conversion of a virtual switch on a host

Some embodiments provide a method for configuring a DCN migrated from a first host computer at a first site of multiple sites spanned by a logical network to a second host computer at a second site. At the second host, the method receives, from the first host, data for configuring the DCN and implementing the logical network on the second site. The data includes (i) data particular to the first site and (ii) a global identifier for a logical port to which the DCN attaches. The method provides the global identifier to a local logical network manager, which uses the global identifier to retrieve data regarding the logical port from a global network manager that manages the logical network across the sites. Based on data retrieved from the global network manager, the method modifies the data particular to the first site into data particular to the second site.

Migration of data compute node across sites

Aspects of the present disclosure relate to flexible and automatic provisioning of virtual instances in a virtual network using single-root input/output virtualization. A device pool is defined that is associated with a virtual entity, such as a virtual switch, in a data center. The device pool identifies available physical hardware devices of one or more host machines, such physical network interfaces connected to the virtual switch, and the device pool is associated with the virtual entity. When a virtual computing instance is connected to the virtual entity, such as when initialized or after migrating between hosts, one or more physical hardware devices are automatically assigned from the device pool to the virtual computing instance based on the association of the device pool to the connected virtual entity. Thus, VMs can be flexibly and automatically provisioned with SR-IOV physical hardware when migrating between hosts using SR-IOV.

Flexible automated provisioning of single-root input/output virtualization (SR-IOV) devices

An approach for an adaptive network input-output control for optimizing allocation of network transmission resources to data flows is provided. In an embodiment, a method comprises: determining, based on, at least in part, default data communications policy, one or more default settings for optimizing allocation of one or more network transmission resources to one or more data flows. The default settings are transmitted to a switch to cause the switch to implement the default settings with respect to the data flows. Upon detecting that stats information about network traffic has been received, one or more updated settings for reallocating at least one of the network transmission resources to at least one of the data flows are determined. The updated settings are transmitted to the switch to cause the switch to implement the updated settings with respect to the at least one of the data flows.

Adaptive network input-output control in virtual environments

A method of configuring a plurality of logical forwarding elements (LFEs) on a physical computing device comprising virtualization software and a plurality of data compute nodes (DCNs) is provided. The method provisions a first LFE configured and managed by a network manager of a datacenter on the physical computing device, the first LFE for performing OSI L3)packet forwarding between DCNs on one or more hosts, the first LFE comprising a plurality of logical ports configured and managed by the network manager, each logical port of the first LFE for connecting a DCN to a logical network configured and managed by the network manager. The method provisions a second LFE configured and managed by a compute manager of the datacenter on the physical computing device, the second LFE for performing OSI L3 packet forwarding between DCNs on one or more hosts, the second LFE comprising a plurality of logical port groups configured and managed by the compute manager, each logical port of the second LFE for connecting a DCN to a logical network configured and managed by the compute manager. The method configures a first DCN in the plurality of DCNs by the compute manager to connect to a logical network configured and managed by the network manager through a port of the first LFE. The method configures a second DCN in the plurality of DCNs by the compute manager to connect to a logical network configured and managed by the compute manager through a port of the second LFE.

Enabling virtual machines access to switches configured by different management entities

Techniques are described for migrating virtual machines (VMs) across virtual switches. To migrate a VM, a destination distributed virtual switch module may, prior to migration, instantiate a distributed virtual port on a destination VM host and reserve the distributed virtual port for each virtual network adapter of the VM. Further, a configuration file for the VM specifying the distributed virtual ports reserved for the virtual network adapters may be copied from a source VM host to the destination VM host or created in the destination VM host. As part of the migration, network state data of the distributed virtual ports used by the VM at the source VM host is copied and applied to the appropriate reserved distributed virtual ports on the destination host. Then, when the migrated VM is powered on at the destination host, the VM configuration specifying the reserved distributed virtual ports is used to attach the virtual network adapters of the VM to the reserved distributed virtual ports.

Live migration of virtual machines across virtual switches in virtual infrastructure

A method of configuring a logical network in a datacenter is provided. The datacenter includes a plurality of host physical computing devices, a compute manager to configure one or more data compute nodes (DCNs) on virtualization software of each host, and a network manager. The method configures, by the network manager, a logical network. the method provides, by the network manager, a read-only configuration construct of the logical network to the virtualization software of a set of hosts in the plurality of hosts. The method obtains, by the compute manager, the read-only configuration construct of the logical network from the virtualization software of the set of hosts. The method configures, by the compute manager, a plurality of DCNs to connect to the logical network using the read only configuration construct of the logical network.

Enabling uniform switch management in virtual infrastructure

A physical computing device is provided that includes a set of processing units and a set of machine readable media (CRM) for execution by the set of processing units. The CRM stores a plurality of data compute nodes (DCNs) deployed by a compute manager of a datacenter, virtualization software, and a network manager agent. The network manager receives a read-only configuration construct for each of a set of logical networks from a network manager of the datacenter, the logical networks configured and managed by the network manager. the agent provisions the set of logical networks on the physical computing device for use by the virtualization software. The agent provides a copy of the read-only construct of each logical network to the compute manager to allow the compute manager to connect one or more DCNs to each of the logical networks.

Method and system of provisioning logical networks on a host machine

A method of migrating a data compute node (DCN) in a datacenter comprising a plurality of host physical computing devices, a compute manager, and a network manager is provided. The method by the network manager configures a plurality of logical networks and provides a read-only configuration construct of at least one of the plurality of logical networks to the virtualization software of each host. The construct of each logical network includes a unique identification of the logical network. The method by the compute manager obtains the unique identification of one or more of the logical networks, determining that a particular logical network is available on first and second hosts based on the unique identification of the particular logical network, and configures a first DCN to move from the first host to the second host based on the determination that the first and second hosts are on the particular logical network.

Growing community of inventors

Mountain View, CA, United States of America

Nilesh Ramchandra Nipane