TECHNIQUES FOR MAINTAINING COMMUNICATIONS SESSIONS AMONG NODES IN A STORAGE CLUSTER SYSTEM

• US 20160062856A1
• Filed: 08/29/2014
• Published: 03/03/2016
• Est. Priority Date: 08/29/2014
• Status: Active Grant

• Current Assignee: NetApp, Inc.
• Original Assignee: NetApp, Inc.
• Inventors: Sundararajan, Manoj, Mu, Paul Yuedong, Ngan, Paul
• Granted Patent: US 9,830,238 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1443 Transmit or communication errors
• CPC Class Codes: G06F 11/2005 using redundant communication controllers
• CPC Class Codes: G06F 11/2071 using a plurality of controllers
• CPC Class Codes: G06F 11/2092 Techniques of failing over between control units
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups

Abstract

Various embodiments are generally directed to techniques for preparing to respond to failures in performing a data access command to modify client device data in a storage cluster system. An apparatus may include a processor component of a first node coupled to a first storage device; an access component to perform a command on the first storage device; a replication component to exchange a replica of the command with the second node via a communications session formed between the first and second nodes to enable at least a partially parallel performance of the command by the first and second nodes; and a multipath component to change a state of the communications session from inactive to active to enable the exchange of the replica based on an indication of a failure within a third node that precludes performance of the command by the third node. Other embodiments are described and claimed.

Claims

1. An apparatus comprising:

a processor component of a first node, the first node coupled to a first storage device storing client device data;

an access component for execution by the processor component to perform a command on the first storage device, the command comprising a data access command to alter the client device data stored in the first storage device and in a second storage device couple to a second node;

a replication component for execution by the processor component to exchange a replica of the command with the second node via a communications session formed between the first and second nodes to enable at least a partially parallel performance of the command by the first and second nodes; and

a multipath component for execution by the processor component to change a state of the communications session from inactive to active to enable the exchange of the replica through the communication session based on an indication of a failure within a third node coupled to the first node that precludes performance of the command by the third node.

2. The apparatus of claim 1, the third node coupled to the first storage device and to the first node as partners of a first high availability (HA) group, the apparatus comprising a partnering component for execution by the processor component to exchange signals with the third node to monitor a status of the third node, detect the failure within the third node and generate the indication of the failure within the third node.

3. The apparatus of claim 2, the partnering component to receive an address of a fourth node from the third node, the multipath component to employ the address to exchange messages with the fourth node to form another communication session between the first and fourth nodes, the fourth node coupled to the second storage device and to the second node as partners of a second HA group.

4. The apparatus of claim 2, the access component to take over control of the first storage device from the third node based on the indication of the failure within the third node.

5. The apparatus of claim 2, comprising:

a memory to store metadata comprising indications of aspects of operation of at least the first node; and

a duplication component to update the metadata with an indication of the first node taking over for the third node in the first HA group, and transmitting a duplicate of the metadata after the update to the third node, the access component to store the metadata after the update within the first storage device.

6. The apparatus of claim 1, the third node coupled to the second storage device and to the second node as partners of a high availability (HA) group, and the replication component to transmit the replica to the third node and receive the indication of failure within the third node from the third node via another communications session formed between the first and third nodes.

7. The apparatus of claim 6, the multipath component to receive an address of the second node from the third node and employ the address to exchange messages with the second node to form the communications session, and the multipath component to form the other communications session with the third node and change a state of the other communications session from active to inactive based on the indication of the failure within the third node.

8. The apparatus of claim 6, comprising:

a memory to store metadata comprising indications of aspects of operation of at least the first node; and

a duplication component to update the metadata with an indication of the second node taking over for the third node in the HA group to receive the replica, and to transmit a duplicate of the metadata after the update to a fourth node, the access component to store the metadata after the update within the first storage device.

9. The apparatus of claim 6, comprising a protocol component coupled to the replication component to receive a request for storage services from a client device via a client interconnect that extends through a network coupled to the first node, convert the request into the command, provide the command to the replication component, and relay an indication of status of performance of the command from the replication component to the client device via the client interconnect.

10. The apparatus of claim 1, the first node belonging to a first high availability (HA) group, the second node belonging to a second HA group, the third node partnered with one of the first node in the first HA group or the second with in the second HA group, and the first, second and third nodes coupled by a inter-cluster interconnect through which the communications session is formed.

11. A computer-implemented method comprising:

performing, at a first node, a command on a first storage device coupled to the first node, the command comprising a data access command to alter client device data stored in the first storage device and in a second storage device coupled to a second node;

exchanging a replica of the command with the second node via a communications session formed between the first and second nodes to enable at least a partially parallel performance of the command by the first and second nodes; and

changing a state of the communications session from inactive to active to enable the exchange of the replica through the communication session based on an indication of a failure within a third node coupled to the first node that precludes performance of the command by the third node.

12. The computer-implemented method of claim 11, the third node coupled to the first storage device and to the first node as partners of a first high availability (HA) group, the method comprising:

exchanging signals with the third node to monitor a status of the third node; and

generating the indication of the failure within the third node based on the signals exchanged with the third node.

13. The computer-implemented method of claim 12, comprising:

receiving an address of a fourth node from the third node; and

employing the address to exchange messages with the fourth node to form another communication session between the first and fourth nodes, the fourth node coupled to the second storage device and to the second node as partners of a second HA group.

14. The computer-implemented method of claim 12, comprising taking over control of the first storage device from the third node based on the indication of the failure within the third node.

15. The computer-implemented method of claim 12, comprising

updating a metadata comprising indications of aspects of operation of at least the first node with an indication of the first node taking over for the third node in the first HA group;

transmitting a duplicate of the metadata after the update to the third node; and

storing the metadata after the update within the first storage device.

16. The computer-implemented method of claim 11, the third node coupled to the second storage device and to the second node as partners of a high availability (HA) group, the method comprising:

transmitting the replica to the third node; and

receiving the indication of failure within the third node from the third node via another communications session formed between the first and third nodes.

17. The computer-implemented method of claim 16, comprising:

receiving an address of the second node from the third node;

employing the address to exchange messages with the second node to form the communications session;

forming the other communications session with the third node; and

changing a state of the other communications session from active to inactive based on the indication of the failure within the third node.

18. The computer-implemented method of claim 16, comprising:

updating the metadata comprising indications of aspects of operation of at least the first node with an indication of the second node taking over for the third node in the HA group to receive the replica;

transmitting a duplicate of the metadata after the update to a fourth node; and

storing the metadata after the update within the first storage device.

19. The computer-implemented method of claim 16, comprising:

receiving a request for storage services from a client device via a client interconnect that extends through a network coupled to the first node;

converting the request into the command; and

transmitting an indication of status of performance of the command to the client device via the client interconnect.

20. The computer-implemented method of claim 11, the first node belonging to a first high availability (HA) group, the second node belonging to a second HA group, the third node partnered with one of the first node in the first HA group or the second with in the second HA group, and the first, second and third nodes coupled by a inter-cluster interconnect through which the communications session is formed.

21. At least one machine-readable storage medium comprising instructions that when executed by a processor component, cause the processor component to:

perform, at a first node, a command on a first storage device coupled to the first node, the command comprising a data access command to alter client device data stored in the first storage device and in a second storage device coupled to a second node;

exchange a replica of the command with the second node via a communications session formed between the first and second nodes to enable at least a partially parallel performance of the command by the first and second nodes; and

change a state of the communications session from inactive to active to enable the exchange of the replica through the communication session based on an indication of a failure within a third node coupled to the first node that precludes performance of the command by the third node.

22. The at least one machine-readable storage medium of claim 21, the third node coupled to the first storage device and to the first node as partners of a first high availability (HA) group, the processor component caused to:

exchange signals with the third node to monitor a status of the third node; and

generate the indication of the failure within the third node based on the signals exchanged with the third node.

23. The at least one machine-readable storage medium of claim 22, the processor component caused to:

receive an address of a fourth node from the third node; and

employ the address to exchange messages with the fourth node to form another communication session between the first and fourth nodes, the fourth node coupled to the second storage device and to the second node as partners of a second HA group.

24. The at least one machine-readable storage medium of claim 22, the processor component caused to take over control of the first storage device from the third node based on the indication of the failure within the third node

25. The at least one machine-readable storage medium of claim 22, the processor component caused to:

update a metadata comprising indications of aspects of operation of at least the first node with an indication of the first node taking over for the third node in the first HA group;

transmit a duplicate of the metadata after the update to the third node; and

store the metadata after the update within the first storage device.

26. The at least one machine-readable storage medium of claim 21, the third node coupled to the second storage device and to the second node as partners of a high availability (HA) group, the processor component caused to:

transmit the replica to the third node; and

receive the indication of failure within the third node from the third node via another communications session formed between the first and third nodes.

27. The at least one machine-readable storage medium of claim 26, the processor component caused to:

receive an address of the second node from the third node;

employ the address to exchange messages with the second node to form the communications session;

form the other communications session with the third node; and

change a state of the other communications session from active to inactive based on the indication of the failure within the third node.

28. The at least one machine-readable storage medium of claim 26, the processor component caused to:

update the metadata comprising indications of aspects of operation of at least the first node with an indication of the second node taking over for the third node in the HA group to receive the replica;

transmit a duplicate of the metadata after the update to a fourth node; and

store the metadata after the update within the first storage device.

29. The at least one machine-readable storage medium of claim 26, the processor component caused to:

receive a request for storage services from a client device via a client interconnect that extends through a network coupled to the first node;

convert the request into the command; and

transmit an indication of status of performance of the command to the client device via the client interconnect.

30. The at least one machine-readable storage medium of claim 21, the first node belonging to a first high availability (HA) group, the second node belonging to a second HA group, the third node partnered with one of the first node in the first HA group or the second with in the second HA group, and the first, second and third nodes coupled by a inter-cluster interconnect through which the communications session is formed.