Hewlett-packard Development Company, L.p.

Hewlett Packard Enterprise Development LP

Balaji Natrajan

Michael G Myrah

Michael Lawrence Sabotta

Sohail Hameed

Pruthviraj Herur Puttaiah

Joseph David Black

Aaron L Jenkins

Lijun Qin

Paul Miller

Keith R Kinnan

Xiaochun Gan

Brian J Corell

Rodrigo Stoll Martins Machado

Greg John Pellegrino

Choon-Seng Tan

Brian M Spencer

Raymond A Ritter

Martin R Rogoff

Charles R Hanna

Schail Hameed

Shilpa B Shrinivas

Chaowen Tan

Henry F Lada

A method for detecting an improper storage system configuration is provided. The method includes receiving, within a first switch, an indication of a configuration change of a storage system. The method also includes performing device discovery to identify devices operatively coupled to the first switch. The method also includes receiving a storage device identifier of a storage device operatively coupled to the first switch. The method also includes searching a repository for the storage device identifier to determine whether the storage device is managed by a second switch.

Storage enclosure bridge detection

Implementations of the disclosed technology may respond to a discovery request according to the origination of the discovery request. In these implementations, a discovery request for a connection to a device assigned to a reusable zone may be obtained. If the discovery request does not originate from a local switch group, the discovery response may indicate that there is no device at the connection.

Reusable zone

Techniques for retrieving data blocks are provided. In one aspect, a storage fabric address of a controller associated with a data block is retrieved by a node. If the node is on the same storage fabric as the retrieved address, the data block may be retrieved over the storage fabric. In another aspect, a directory server maintains mappings of data blocks to storage fabric addresses of controllers associated with the data blocks. A request for the location of the data block includes the storage fabric address of the associated controller.

Storage fabric address based data block retrieval

A hardware device detects change messages broadcast within a system. The system includes the hardware device, one or more controller devices, one or more expander devices, and one or more target devices interconnected among one another. The hardware device determines whether the change messages were broadcast within the system every first period of time or less for at least a second period of time, the first period of time less than the second period of time. In response to determining that the change messages were broadcast within the system every first period of time or less for at least the second period of time, the hardware devices signals that an error has been detected.

Change message broadcast error detection

A serial attached small computer system interface (SAS) expander for providing a virtual tape drive. The expander includes a virtual tape manager to generate a virtual tape virtual PHY capable of being part of a zone group wherein the virtual tape virtual PHY comprises drive bays which support disk drives for emulation of tape drives.

SAS virtual tape drive

In a method to zone data to a virtual machine, support information is exchanged between a physical controller and an expander, the support information comprising an indication of an ability of the physical controller and the expander to support thereon a set of virtual controllers and a set of virtual physical layers (virtual PHYs), respectively. In response to a positive indication sending a list of SAS addresses to the expander, a list of SAS addresses is sent to the expander. Then, a first SAS address of the list of SAS addresses is assigned to a first virtual PHY of the set of virtual PHYs, wherein the assigning is performed by the expander, and wherein the first virtual PHY at an assigned SAS address corresponds to a first virtual controller of the set of virtual controllers.

Zoning data to a virtual machine

Techniques for operating a serial attached SCSI (SAS) expander that includes a latency table comprising entries of outbound PHYS with latency values associated with connections between inbound PHYs and outbound PHYs. A storage management module to, in response to receipt of a command from an initiator device associated with an inbound PHY to route data to a target device associated with an outbound PHY, select from the latency table a random outbound PHY from among a plurality of outbound PHYs, wherein the random selection is based on weighted average of latency values of the outbound PHY entries of the latency table.

SAS latency based routing

A method includes detecting a topology change on a link in a network and determining which zone group corresponds to the topology change. The method also includes identifying a zone group in the network that is permitted to access a device on said link and determining an address of an expander that has the zone group permitted to access the device on the link. The method further includes generating an expander-targeted zoned broadcast command to report a topology change in the network. The expander-targeted zoned broadcast command has, as a destination address, the address of the expander determined to have a zone group permitted to access a device on the link.

Expander-targeted zoned broadcast

In one example in accordance with aspects of the present disclosure, an expander is provided. The expander comprises a workload scheduling module to cause the expander to enter a first mode of operation where the expander processes interrupts, and further to enter a second mode of operation where the expander processes interrupts for up to a predetermined time period before responding to at least one of Serial Management Protocol (SMP) commands and Serial SCSI Protocol (SSP) commands with a retry message.

Expander interrupt processing

An example method for memory resource provisioning using SAS zoning can include a serial attached Small Computer System Interface (SCSI) (SAS) switch and a provisioning manager (PM) executing on the SAS switch. A plurality of servers are communicatively coupled to the SAS switch, each server executing an operating system (OS) having access to one or more OS volumes. A host provisioning agent (HPA) executes on the OS of each of the plurality of servers. A plurality of memory resources are communicatively coupled to the SAS switch. Each memory resource can have a number of physical hard drives. A first portion of the memory resources is associated with zone groups corresponding to the plurality of servers. A second portion of the memory resources is associated with a provisioning zone group (PZG) hidden from the plurality of servers but available to the PM.

Memory resource provisioning using SAS zoning

An example method includes (i) creating, by a first serial attached SCSI (SAS) switch, a first topology map describing a portion of a SAS fabric associated with the first SAS switch; (ii) receiving, at the first SAS switch and from a second SAS switch, a second topology map describing a portion of the SAS fabric associated with the second SAS switch; and (iii) merging, by the first SAS switch, the first topology map and the second topology map to produce a consolidated topology map of the SAS fabric.

SAS fabric discovery

A method and apparatus for assigning zone groups to a storage enclosure is disclosed. When a storage enclosure is added to a switch in a fabric, a signature of the storage enclosure will be created from the storage devices loaded in that enclosure. The signature will then be compared against the signature from a storage enclosure that is offline. When the signature matches the signature of the offline storage enclosure, the zone groups from the offline storage enclosure will be copied to the storage enclosure that was added to the fabric.

Growing community of inventors

Spring, TX, United States of America

Balaji Natrajan