Microsoft Technology Licensing, LLC

Shekhar Agarwal

Deepak Bansal

Geoffrey Hugh Outhred

Narasimhan Agrahara Venkataramaiah

Rishabh Tewari

Tomas Talius

Harish Kumar Chandrappa

Parag Sharma

Selim Ciraci

Nimish Aggarwal

Anitha Adusumilli

Longzhang Fu

Dhruv Malik

Joemmanuel Ponce Galindo

The techniques disclosed herein improve the efficiency, reliability and scalability of flow processing systems by providing a multi-tier flow cache structure that can reduce the size of a flow table and also reduce replicated flow sets. In some configurations, a system can partition a flow space across workers and replicate the flows within a partition to a set of workers. In some configurations, a flow cache structure can include three tiers: (1) a scalable flow processing layer for executing the actions and transformations of a flow, (2) a flow state management layer for managing distributed flow state decisions, and (3) a flow decider layer for identifying actions and transformations needs to be executed on each packet of a flow. Flow replications allow other workers to pick up flows allocated to a particular worker that is taken offline in the event of a crash or update.

Distributed flow processing and flow cache

A load balancer capable of instantiating a data plane within the load balancer, deleting the data plane from the load balancer, and/or enacting a change to the data plane. The load balancer instantiates a data plane for an identified tenant. The instantiated data plane is placed in a data path of network data transmitted from one or more sources to a plurality of tenant addresses that each corresponds to a different tenant or group of tenants. The instantiated data plane is also dedicated to the identified tenant such that the data plane isolates first network data destined to a first tenant address that corresponds to the identified tenant from second network data destined to one or more other tenant addresses. The load balancer also deletes the instantiated data plane from the load balancer, or enacts a change to the instantiated data plane.

Data plane manipulation in a load balancer

Techniques for allowing access to shared cloud resource using private network addresses are disclosed herein. In one embodiment, a connection packet representing a connection request to a shared cloud resource in the cloud computing system can be intercepted. In response, the connection packet can be encapsulated with data representing one or more of a VNET ID, a VNET source address, or a VNET destination address of a virtual network from which the connection packet is received. The encapsulated connection packet can then be forwarded to the shared cloud resource while retaining the data representing one or more of the VNET ID, the VNET source address, or the VNET destination address for access control at the shared cloud resource.

Accessing cloud resources using private network addresses

A load balancer capable of adjusting how network data is distributed to a tenant or group of tenants by manipulating the data plane. The load balancer is placed directly in the flow path of network data that is destined for a tenant or group of tenants having a tenant address. The load balancer includes a control plane and one or more data planes. Each data plane may contain one or more network traffic multiplexors. Further, each data plane may be dedicated to a tenant or group of tenants. Data planes may be added or deleted from the load balancer; additionally, multiplexors may be added or deleted from a data plane. Accordingly, network data directed towards one tenant is less likely to affect the performance of load balancing performed for another tenant.

Growing community of inventors

Bellevue, WA, United States of America

Shekhar Agarwal