Storage system transactions

Patent metadata

Field	Value
Publication number	US-2016350192-A1
Application number	US-201415114900-A
Country	US
Kind code	A1
Filing date	Mar 20, 2014
Priority date	Mar 20, 2014
Publication date	Dec 1, 2016
Grant date	—

Abstract

A system that includes a storage volume comprising a plurality of regions. The system also includes a plurality of nodes to receive transactions related to the storage volume from one or more client computers and execute the transactions. Each node is designated as an owner of one region of the plurality of regions. Each one of the plurality of nodes includes a ticket dispenser for dispensing tickets that ensure that the transactions that are actively being executed have exclusive access to a storage object identified in the transaction.

First claim

What is claimed is:

1 . A system comprising:
a storage volume comprising a plurality of regions; a plurality of nodes to receive transactions related to the storage volume from one or more client computers and execute the transactions, each node designated as an owner of one region of the plurality of regions; and wherein each one of the plurality of nodes includes a ticket dispenser for dispensing tickets that ensure that the transactions that are actively being executed have exclusive access to a storage object identified in the transaction.

2 . The system of claim 1 , wherein only the owner of the region can execute transactions affecting the region.

3 . The system of claim 1 , when the ticket dispenser issues each ticket, the ticket is stored in a transaction log that is used to check the transactions that are actively being executed when a new transaction is received.

4 . The system of claim 1 , when the ticket dispenser issues each ticket, the ticket dispenser sends the ticket to another node of the plurality of nodes that has been designated as a backup node for the transaction.

5 . The system of claim 4 , wherein the backup node is identified the node that would take ownership of the storage object identified in the transaction if the node that generated the ticket fails.

6 . A method comprising:
receiving a new transaction at a plurality of nodes of a storage system from a client computer; transferring the new transaction to an owner node of the plurality of nodes, the owner node controlling a region of the storage volume that is affected by the transaction; and at the owner node:
checking a transaction log of the owner node for active transactions being executed by the owner node; and
if the new transaction does not conflict with any of the active transactions, issuing a ticket for the new transaction, wherein the ticket ensures exclusive access to a storage object that is affected by the transaction.

7 . The method of claim 6 , comprising storing the ticket to the transaction log of the owner node.

8 . The method of claim 6 , comprising sending the ticket to a backup node of the plurality of nodes.

9 . The method of claim 8 , comprising selecting the backup node as a node of the plurality of nodes that is designated to become the new owner node of the storage object if the owner node fails.

10 . The method of claim 8 , comprising, at the owner node, receiving a backup transaction from a second node of the plurality of nodes and storing the backup transaction to a backup transaction log.

11 . The method of claim 10 , comprising:
detecting failure of the second node; taking ownership one of the storage objects associated with the second node; and executing the backup transaction.

12 . A tangible, non-transitory, computer-readable medium comprising instructions that direct a processor to:
receive a new transaction from a client computer at a first node of a storage system; if a region affected by the new transaction is owned by a second node, send the transaction to the second node for execution by the second node; and if the region affected by the new transaction is owned by the first node:
check a transaction log of the first node for active transactions being executed by the first node;
if the new transaction does not conflict with any of the active transactions, issuing a ticket for the new transaction, wherein the ticket ensures exclusive access to a storage object that is affected by the transaction; and
execute the transaction.

13 . The computer-readable medium of claim 12 , comprising instructions that direct the processor to store the ticket to the transaction log of the first node, and send the ticket to a backup node of the plurality of nodes.

14 . The computer-readable medium of claim 13 , comprising instructions that direct the processor to select the backup node based on the backup node being designated to become the new owner node of the storage object if the first node fails.

15 . The computer-readable medium of claim 12 , comprising instructions that direct the processor to receiving a backup transaction from the second node of the plurality of nodes and store the backup transaction to a backup transaction log.

Assignees

• Hewlett Packard Entpr Dev Lp

Inventors

• Doherty Mark
• Nazari Siamak
• Stewart Jonathan
• Dalzell Richard
• Hynes Peter

Classifications

• G06F3/065

  Replication mechanisms · CPC title

• G06F11/2025

  using centralised failover control functionality · CPC title

• G06F13/385 Primary

  for adaptation of a particular data processing system to different peripheral devices · CPC title

• G06F11/2092 Primary

  Techniques of failing over between control units · CPC title

• G06F3/0665

  at area level, e.g. provisioning of virtual or logical volumes · CPC title