System and method for facilitating protection against run-away subnet manager instances in a middleware machine environment

Patent metadata

Field	Value
Publication number	US-9455898-B2
Application number	US-201113235144-A
Country	US
Kind code	B2
Filing date	Sep 16, 2011
Priority date	Sep 17, 2010
Publication date	Sep 27, 2016
Grant date	Sep 27, 2016

Abstract

A system and method can support a middleware machine environment that includes a set of subnet manager instances, which reside on one or more nodes in the middleware machine environment and cooperate to provide a highly available subnet manager service within a subnet, wherein each said subnet manager instance is associated with a different private secure key. The subnet manager instances can negotiate with each other and elect a master subnet manager responsible for configuring and managing the middleware machine environment using the private secure key associated with the master subnet manager. The subnet can be reconfigured to be associated with a different private secure key, when a new subnet manager instance is elected as the master subnet manager. An old master subnet manager can be automatically prevented from resuming normal operations as the master subnet manager, in order to avoid undesired consequence such as a “split brain” scenario.

First claim

What is claimed is:

1. A system for supporting a middleware machine environment, comprising:
one or more microprocessors;
one or more network switches, wherein each said network switch provides one or more ports that are used to connect with servers in a middleware machine environment, and wherein the one or more network switches are configured to provide a subnet;
a set of subnet manager instances that resides on one or more nodes running on the one or more microprocessors in the middleware machine environment, wherein the set of subnet manager instances cooperate to provide a subnet manager service within the subnet, wherein each said subnet manager instance is associated with a different key value or range that is known to other subnet manager instances, and
wherein the set of subnet manager instances can negotiate with each other and elect a master subnet manager, which is responsible for configuring and managing the middleware machine environment using the key value or range that is associated with the master subnet manager, including that
the set of subnet manager instances determine which key value or range is currently in use; and based on that determination determine whether to elect a master subnet manager.

2. The system according to claim 1 , further comprising a separate storage system that connects with the plurality of host servers through said one or more network switches.

3. The system according to claim 1 , wherein: the subnet is an Infiniband (IB) subnet.

4. The system according to claim 1 , further comprising: one or more gateway instances that can be accessed by a guest.

5. The system according to claim 1 , wherein: the subnet manager instances can communicate with each other using an in-band communication protocol.

6. The system according to claim 1 , wherein: the subnet can be divided into a dynamic set of resource domains implemented by subnet partitions.

7. The system according to claim 1 , wherein: the master subnet manager can use a default partitioning policy for initialization when no partitioning policy is specified.

8. The system according to claim 1 , wherein: the key value is a M_Key that is a 64 bit secret value that is known only to authorized entities in the subnet.

9. The system according to claim 8 , wherein: when a subnet management agent (SMA) associated with a port in the subnet is configured with a M_Key value, an in-band request needs to specify the M_Key value in order to change a state associated with the port.

10. The system according to claim 1 , wherein: the subnet is reconfigured to be associated with a different key value or range, when a different subnet manager instance is elected as the master subnet manager.

11. The system according to claim 1 , wherein: each different key value or range is defined in a different range that is known to other subnet manager instances in the subnet.

12. The system according to claim 1 , wherein: the set of subnet manager instances in the subnet can dynamically determine which key value or range in a defined range is in use depending on which subnet manager instance is currently the master subnet manager.

13. The system according to claim 1 , wherein: an old master subnet manager is automatically prevented from resuming normal operations as a master subnet manager after a new master subnet manager is elected in order to prevent a split brain scenario.

14. The system according to claim 1 , wherein: one or more updated subnet configuration policies can be applied in the subnet through one or more subnet managers that can recognize the key value or range associated with the updated subnet configuration policies, and other subnet managers in the subnet that cannot recognize the key value or range can be left unaffected by the one or more updated subnet configuration policies and be prevented from updating a state of the subnet.

15. The system according to claim 1 , wherein: the master subnet manager can determine that a connection to a new node is unintentional since the master subnet manager cannot recognize the private secure key used for the node.

16. The system according to claim 15 , wherein: the master subnet manager in a remote subnet containing the new node is not allowed to change a state in the subnet and the master subnet manager in the subnet is not allowed to change a state in the remote subnet.

17. A method for supporting a middleware machine environment, comprising:
providing one or more network switches, wherein each said network switch provides one or more ports that are used to connect with servers in a middleware machine environment, and wherein the one or more network switches are configured to provide a subnet;
associating a different key value or range with each subnet manager instance in a set of subnet manager instances, wherein each said subnet manager instance is associated with different key values or ranges that is known to other subnet manager instances and each said subnet manager instance resides on one or more nodes running on one or more microprocessors in the middleware machine environment, wherein the set of subnet manager instances cooperate to provide a subnet manager service within the subnet;
allowing the set of subnet manager instances to negotiate with each other and elect a master subnet manager; and
configuring and managing the middleware machine environment using the private secure key that is associated with the master subnet manager, including that
the set of subnet manager instances determine which key value is currently in use; and, based on that determination determine whether to elect a master subnet manager.

18. A non-transitory machine readable medium having instructions stored thereon that when executed cause a system including one or more network switches, wherein each said network switch provides one or more ports that are used to connect servers in a middleware machine environment, and wherein the one or more network switches are configured to provide a subnet to perform the steps of:
associating a different key value or range with each subnet manager instance in a set of subnet manager instances, wherein each said subnet manager instance is associated with different key values or ranges that is known to other subnet manager instances and each said subnet manager instance resides on one or more nodes running on one or more microprocessors in a middleware machine environment, wherein the set of subnet manager instances cooperate to provide a subnet manager service within the subnet;
allowing the set of subnet manager instances to negotiate with each other and elect a master subnet manager; and
configuring and managing the middleware machine environment using the private secure key that is associated with the master subnet manager, including that
the set of subnet manager instances determine which key value is currently in use; and, based on that determination determine whether to elect a master subnet manager.

Assignees

• Johnsen Bjorn-Dag
• Holen Line
• Moxnes Dag Georg
• Oracle Int Corp

Inventors

• Johnsen Bjorn-Dag
• Holen Line
• Moxnes Dag Georg

Classifications

• H04L12/462

  LAN interconnection over a bridge based backbone · CPC title

• H04L69/40

  for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection (management of faults, events, alarms or notifications in data switching networks H04L41/06) · CPC title

• H04L45/50

  using label swapping, e.g. multi-protocol label switch [MPLS] · CPC title

• H04L45/00 Primary

  Routing or path finding of packets in data switching networks (routing or path finding in wireless networks H04W40/00) · CPC title

• H04L41/40

  using virtualisation of network functions or resources, e.g. SDN or NFV entities · CPC title