Brocade Communications Systems, Inc.

Avago Technologies International Sales Pte. Limited

Brocade Communication Systems LLC

Balakumar N Kaushik

Richard L Hammons

Vincent W Guan

Shankar Balasubramanian

Vineet M Abraham

Christopher A Del Signore

Santosh S Shanbhag

Pranab Patnaik

Sathish Kumar Gnanasekaran

Satish K Gnanasekaran

Wai-Sze Tam

An interfabric link between two separate Fibre Channel fabrics so that devices in one fabric can communicate with devices in another fabric without requiring the merger of the two fabrics. The interfabric switch performs a conversion or a translation of device addresses in each fabric so that they are accessible to the other fabric. This translation is preferably done using a private to public loop address translation. In a first embodiment the external ports of the interfabric switch are configured as E_ports. A series of internal ports in each interfabric switch are joined together forming a series of virtual or logical switches. The virtual switches are then interconnected using private loops. The use of the private loop is enabled by the presence of translation logic which converts fabric addresses to loop addresses and back so that loop and fabric devices can communicate. Because each port can do this translation and the private loop addressing does not include domain or area information, the change in addresses between the fabrics is simplified. In a second embodiment the external ports are configured as NL_ports and the connections between the virtual switches are E_ports. Thus the private to public and public to private translations are done at the external ports rather than the internal ports as in the prior embodiment. The virtual switches in the interfabric switch match domains with their external counterparts so that the virtual switches effectively form their own fabric, connected to the other fabrics by the private loops.

Method and apparatus for routing between fibre channel fabrics

An interfabric link between two separate Fiber Channel fabrics so that devices in one fabric can communicate with devices in another fabric without requiring the merger of the two fabrics. The interfabric switch performs a conversion or a translation of device addresses in each fabric so that they are accessible to the other fabric. This translation is preferably done using a private to public loop address translation. In a first embodiment the external ports of the interfabric switch are configured as E_ports. A series of internal ports in each interfabric switch are joined together forming a series of virtual or logical switches. The virtual switches are then interconnected using private loops. The use of the private loop is enabled by the presence of translation logic which converts fabric addresses to loop addresses and back so that loop and fabric devices can communicate. Because each port can do this translation and the private loop addressing does not include domain or area information, the change in addresses between the fabrics is simplified. In a second embodiment the external ports are configured as NL_ports and the connections between the virtual switches are E_ports. Thus the private to public and public to private translations are done at the external ports rather than the internal ports as in the prior embodiment. The virtual switches in the interfabric switch match domains with their external counterparts so that the virtual switches effectively form their own fabric, connected to the other fabrics by the private loops.

Copy capability moved into the SAN fabric and provided as a data mover service. A well-known address is utilized to receive copy commands from the hosts. Each switch in the fabric contains a front end or service interface which receives the copy command and manages access control or zoning and LUN mapping. LUN mapping and zoning are based on shared databases. Assuming correct zoning, the service interface will perform any necessary LUN mapping on the received copy command and forward the command to a copy engine to perform the copy operation. The switch service interface also maintains copy operation status, querying the copy engine if necessary, so that any host status requests are also simply handled. Additionally, if multiple copy engines are present, the switches can communicate to perform load balancing.

Fibre channel fabric copy service

An interfabric link between two separate Fiber Channel fabrics so that devices in one fabric can communicate with devices in another fabric without requiring the merger of the two fabrics. The interfabric switch performs a conversion or a translation of device addresses in each fabric so that they are accessible to the other fabric. In a first embodiment the external ports of the interfabric switch are configured as E_ports. A series of internal ports in each interfabric switch are joined together forming a series of virtual or logical switches. In a second embodiment the external ports are configured as NL_ports and the connections between the virtual switches are E_ports. The virtual switches in the interfabric switch match domains with their external counterparts so that the virtual switches effectively form their own fabric.

An isolation switch blade Fibre Channel switch presents F_ports to form a first Fibre Channel fabric and N_ports to a second Fibre Channel fabric to appear as node devices. The isolation switch blade may be used to connect a plurality of blade servers to a Fibre Channel fabric. Fabric events engendered by the insertion or removal of hot-pluggable devices are handled by the isolation switch blade and 'event storms' on the Fibre Channel fabric are avoided. The isolation switch blade presents the blade servers to the FC fabric as a virtualized N_port.

Isolation switch for fibre channel fabrics in storage area networks

An isolation switch blade Fiber Channel switch presents F_ports to form a first Fiber Channel fabric and N_ports to a second Fiber Channel fabric to appear as node devices. The isolation switch blade may be used to connect a plurality of blade servers to a Fiber Channel fabric. Fabric events engendered by the insertion or removal of hot-pluggable devices are handled by the isolation switch blade and 'event storms' on the Fiber Channel fabric are avoided. The isolation switch blade presents the blade servers to the FC fabric as a virtualized N_port.

The snapshot capability moving into the SAN fabric and being provided as a snapshot service. A well-known address is utilized to receive snapshot commands. Each switch in the fabric connected to a host contains a front end or service interface to receive the snapshot command. Each switch of the fabric connected to a storage device used in the snapshot process contains a write interceptor module which cooperates with hardware in the switch to capture any write operations which would occur to the snapshot data area. The write interceptor then holds these particular write operations until the original blocks are transferred to a snapshot or separate area so that the original read data is maintained. Should a read operation occur to the snapshot device and the original data from requested location has been relocated, a snapshot server captures these commands and redirects the read operation to occur from the snapshot area. If, however, the read operation is directed to the original drive, the read is provided from the original data areas, even if the data had been replaced. The snapshot server determines the existence of particular snapshot devices, allocates their storage locations, provides this information to both the service interfaces and the write interceptors and handles read and write operations to the snapshot device.

Fibre channel fabric snapshot server

Write capture for fibre channel fabric snapshot service

Service interface for fibre channel fabric snapshot service

Network with fibre channel fabric snapshot service

Fault tolerant techniques to update a distributed database in a switched-fabric network. A commit master switch initiates an update operation using a four-phase process that provides for the automatic fail-over to a second (or more) switches if the commit master should not complete the update operation. In response to a lost commit master switch, one or more switches may assume the role of the commit master to either abort or complete the update operation.

Growing community of inventors

Sunnyvale, CA, United States of America

Balakumar N Kaushik