Data storage network with host transparent failover controlled by host bus adapter

• US 20030126315A1
• Filed: 12/28/2001
• Published: 07/03/2003
• Est. Priority Date: 12/28/2001
• Status: Active Grant

• Current Assignee: Valtrus Innovations Limited (f/k/a Dolya Holdco 9 Limited) (Key Patent Innovations Limited)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: White, Michael D., Tan, Choon-Seng, Pellegrino, Greg J.
• Granted Patent: US 7,111,084 B2
• Time in Patent Office: Days
• US Class Current: 710/1
• CPC Class Codes: G06F 11/201 between storage system components
• CPC Class Codes: G06F 11/2025 using centralised failover control functionality
• CPC Class Codes: G06F 11/2092 Techniques of failing over between control units
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]

Abstract

A method and host bus adapter for controlling failover and failback processes within a redundant data storage system. The method is carried out by a failover mechanism in a host bus adapter linked to a host server. The system includes a communication fabric providing data paths to pairs of redundant storage controllers managing access to storage devices (such as hard disk drives arranged as RAID). The method includes operating the host bus adapter to detect a failover condition and in response, to match the failover condition to a particular failover action. The failover conditions and corresponding actions are stored in a well-defined rule set accessed by the host bus adapter. The method continues with performing with the host bus adapter the matched failover action. The method is transparent to the host computer device, as the host is not required to participate in the failover processes.

Claims

1. A method of controlling a failover process in a data storage system including a host, a host bus adapter, a communication fabric including data paths, and standby and active storage controllers, comprising:

detecting with the host bus adapter a failover condition;

responsive to the detecting, operating the host bus adapter to match the failover condition to a particular failover action in a failover rule set; and

performing with the host bus adapter the matched failover action.

2. The method of claim 1, wherein the detecting, operating, and the failover action performing are completed without acts initiated by the host.

3. The method of claim 2, wherein the detecting includes identifying a particular failure type and wherein the particular failover action is selected from an action subset corresponding to the particular failure type.

4. The method of claim 3, wherein the failure type is selected from the group consisting of inter-controller link down, the active storage controller failed, the standby controller failed, an active path failed, and a standby path failed.

5. The method of claim 1, further including prior to the performing, determining with the host bus adapter if all active paths have failed and if all active paths determined failed, skipping the failover action performing when the host bus adapter determines either all other available paths have failed or a standby path is marked as unusable.

6. The method of claim 1, further including after the failover action performing, operating the host bus adapter to initiate failback when a controller in a preferred slot is replaced, when the controller in the preferred slot is rebooted, and when unusable paths become usable.

7. The method of claim 1, further including performing load distribution with the host bus adapter between the host and the controllers.

8. The method of claim 1, further including enforcing with the host bus adapter anti-thrashing rules comprising preventing the performing from being completed more than set number of times per pre-set monitoring interval.

9. A host bus adapter for managing failover and failback processes within a data storage system having a host server, a communication fabric, at least one active storage controller, and at least one standby storage controller, comprising:

a connector linking the host bus adapter to a processor of the host server;

a port linking the host bus adapter to the communication fabric configured for transmitting and receiving digital information; and

a failover mechanism detecting a redundancy failure in the data storage system and in response, initiating failover actions.

10. The host bus adapter of claim 9, wherein the failover actions are selected by the failover mechanism from a failover rule set.

11. The host bus adapter of claim 10, wherein the failover mechanism is further configured to determine at the time of the detecting, operating conditions within the data storage system, to determine whether the operating conditions match a set of failover conditions, and if matching, to select the failover action corresponding to the operating conditions.

12. The host bus adapter of claim 11, wherein the failover conditions are specific to the detected redundancy failure.

13. The controller of claim 9, wherein the failover mechanism presents a single logical unit number (LUN) entity to operating system device drivers in the host processor that is discoverable a plurality of times and wherein the failover actions are initiated without prior communication with the host processor.

14. A data storage system with redundant data storage, comprising:

a host computer device with a processor running operating system device drivers;

a communication fabric for carrying digital data signals;

an active controller controlling access by the host computer device to data storage devices;

a standby controller controlling access by the host computer device to the data storage devices; and

a host bus adapter linked to the host processor and the communication fabric for selecting a path through the communication fabric to one of the active and standby controllers for providing the operating system device drivers with access to the data storage devices, wherein host bus adapter is configured to initiate a failover action selected from a set of failover actions.

15. The system of claim 14, wherein the host bus adapter detects a potential failure in redundancy and determines whether to initiate the failover action by determining whether failover operating circumstances and failover operating conditions for the potential failure are satisfied.

16. The system of claim 15, wherein the failover operating circumstances require when an active path in the communication fabric fails that at least one path to the controllers is available and that a path to the standby controller is usable.

17. The system of claim 15, wherein the initiated failover action is selected from the set of failover actions based on existing ones of the failover operating conditions.

18. The system of claim 14, wherein the data storage devices are grouped into subsets and wherein the host bus adapter is configured to perform the failover action for the subsets when a particular storage device within the subset requires the failover action.

19. The system of claim 14, wherein the host bus adapter is adapted to initiate failover actions for a particular one or group of the data storage devices less than a preset number of times per monitoring interval.

20. The system of claim 14, wherein the host bus adapter presents a single logical unit number (LUN) entity to each of the operating system device drivers that is discoverable multiple times.