System and method for supporting well defined subnet topology in a middleware machine environment

Patent metadata

Field	Value
Publication number	US-10630570-B2
Application number	US-201113235161-A
Country	US
Kind code	B2
Filing date	Sep 16, 2011
Priority date	Sep 17, 2010
Publication date	Apr 21, 2020
Grant date	Apr 21, 2020

Abstract

A system and method can support a middleware machine environment. The middleware machine environment can include a subnet manager and a plurality of fabric components, wherein each said fabric component is associated with a subnet management agent (SMA). The subnet manager maintains a topology graph that indicates how ports are connected in the subnet. The subnet manager operates to periodically reevaluate one or more states associated with each fabric component in the subnet by determining whether each SMA is responding correctly to a request.

First claim

What is claimed is:

1. A method for supporting a middleware machine environment, the middleware machine environment operating, at least partially in a subnet of a network fabric, the subnet including a plurality of fabric components including a plurality of end node servers and plurality of network switches, wherein each fabric component in said plurality of fabric components comprises a subnet management agent (SMA), said method comprising:
providing a network switch comprising one or more microprocessors;
providing a subnet manager running on said network switch;
discovering the topology of the subnet, including a link from a particular fabric component of the plurality of fabric components to the subnet manager;
maintaining, with said subnet manager, a subnet connectivity state represented by a topology graph of the subnet consisting of a list of links connecting fabric components of said plurality of fabric components included in the subnet, wherein the list of links includes the link from the particular fabric component to the subnet manager, wherein each of said links identifies a pair of ports connecting fabric components of said plurality of fabric components included in the subnet, and wherein said subnet connectivity state is maintained by:
transmitting a subnet management packet from the subnet manager to the particular fabric component,
periodically determining whether the link from the particular fabric component of the plurality of fabric components to the subnet manager is up, and
modifying said topology graph of the subnet to generate a modified topology graph of the subnet by:
modifying, in said list of links in the topology graph of the subnet, the link from the particular fabric component to the subnet manager to indicate that said link is down when the link from the particular fabric component to the subnet manager has an observed physical state of down, and
modifying, in said list of links in the topology graph of the subnet, the link from the particular fabric component to the subnet manager to indicate that said link is down when the link from the particular fabric component to the subnet manager has an observed physical state of up but the particular fabric component has an SMA which does not respond correctly to said subnet management packet; and

establishing, with said subnet manager, routes and at least one alternative route for routing packets between all of said plurality of end node servers within said subnet and for the particular fabric component using a routing algorithm, wherein said routing algorithm is implemented using the subnet connectivity state defined in said modified topology graph of the subnet.

2. The method of claim 1 , further comprising:
assigning, by the subnet manager, a common subnet ID and address to each port in the subnet.

3. The method of claim 1 , further comprising:
sweeping, by the subnet manager, the subnet and looking for topology updates in the subnet.

4. The method of claim 1 , further comprising:
using a routing algorithm that does not take into account the various states associated with said SMAs.

5. The method of claim 1 , wherein:
the subnet manager is a master subnet manager that manages said plurality of fabric components included in the subnet.

6. The method of claim 1 , further comprising:
implementing said routing algorithm independently of the logical state of the subnet management agents associated with each of the plurality of fabric components included in the subnet.

7. The method of claim 1 , further comprising:
providing the subnet with a separate storage system.

8. A system for supporting a middleware machine environment, the middleware machine environment operating, at least partially in a subnet of a network fabric, the subnet including a plurality of fabric components including a plurality of end node servers and a plurality of network switches, wherein each fabric component in said plurality of fabric components comprises a subnet management agent (SMA), said system comprising:
a network switch comprising one or more microprocessors;
a subnet manager running on the one or more microprocessors of said network switch;
wherein the subnet manager is configured to:
discover the topology of the subnet, including a link from a particular fabric component of the plurality of fabric components to the subnet manager;
maintain a subnet connectivity state represented by a topology graph of the subnet consisting of a list of links connecting fabric components of said plurality of fabric components included in the subnet, wherein the list of links includes the link from the particular fabric component to the subnet manager, wherein each of said links identifies a pair of ports connecting fabric components of said plurality of fabric components included in the subnet, and wherein said subnet connectivity state is maintained by:
transmitting a subnet management packet from the subnet manager to the particular fabric component,
periodically determining whether the link from the particular fabric component of the plurality of fabric components to the subnet manager is up, and
modifying said topology graph of the subnet to generate a modified topology graph of the subnet by:
modifying, in said list of links in the topology graph of the subnet, the link from the particular fabric component to the subnet manager to indicate that said link is down when the link from the particular fabric component to the subnet manager has an observed physical state of down, and
modifying, in said list of links in the topology graph of the subnet, the link from the particular fabric component to the subnet manager to indicate that said link is down when the link from the particular fabric component to the subnet manager has an observed physical state of up but the particular fabric component has an SMA which does not respond correctly to said subnet management packet; and

wherein the subnet manager is further configured to establish routes and at least one alternative route for routing packets between all of said plurality of end node servers within said subnet and for the particular fabric component using a routing algorithm, wherein said routing algorithm is implemented using the subnet connectivity state defined in said modified topology graph of the subnet.

9. The system according to claim 8 , wherein:
the subnet is in a fat-tree topology.

10. The system according to claim 8 , wherein:
the subnet manager assigns a common subnet ID and address to each port in the subnet.

11. The system according to claim 8 , wherein:
the subnet manager sweeps the subnet and looks for topology update in the subnet.

12. The system according to claim 8 , wherein:
the subnet manager sends subnet management request packets (SMPs) to a SMA on the particular fabric component in the subnet via a subnet management interface (SMI) on the particular fabric component.

13. The system according to claim 12 , wherein:
the SMA operates to return a response SMP to the SMI, which in turn transmits the response SMP packet back to the subnet manager.

14. The system according to claim 8 , wherein:
the topology graph of the subnet allows the subnet to use a routing algorithm that does not take into account the various states associated with said SMAs.

15. The system according to claim 8 , further comprising:
the subnet manager is a master subnet manager that manages said plurality of fabric components included in the subnet.

16. The system according to claim 8 , wherein:
the subnet includes a separate storage system.

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

• H04L41/0659

  by isolating or reconfiguring faulty entities · CPC title