WO2012037512A1

WO2012037512A1 - Performing partial subnet initialization in a middleware machine environment

Info

Publication number: WO2012037512A1
Application number: PCT/US2011/052021
Authority: WO
Inventors: Bjorn-Dag Johnsen; Line Holen; Roy Arntsen
Original assignee: Oracle International Corporation
Priority date: 2010-09-17
Filing date: 2011-09-16
Publication date: 2012-03-22
Also published as: US20120072563A1; JP5844373B2; WO2012037520A1; JP5885747B2; WO2012037518A1; CN103125102B; EP2617157A1; EP2617157B1; US20120072562A1; EP2617165A1; EP2617159A1; CN103125102A; US20120069730A1; US20120079580A1; CN103125097B; JP2013543304A; US20120072564A1; JP5893628B2; CN103125097A; US20120079090A1

Abstract

A system and method can perform subnet initialization in a middleware machine environment. In accordance with one embodiment, a plurality of management nodes in the middleware machine environment can be interconnected so that a plurality of subnet managers that are associated with the plurality of management nodes can communicate with each other using an in-band communication protocol. The plurality of subnet managers can negotiate and elect a master subnet manager, which operates to configure and manage the middleware machine environment. The master subnet manager can replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol. The master subnet manager can then set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

Description

PERFORMING PARTIAL SUBNET INITIALIZATION IN A MIDDLEWARE MACHINE

ENVIRONMENT

COPYRIGHT NOTICE

A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.

Field of Invention:

[0001] The present invention is generally related to computer systems and software such as middleware, and is particularly related to supporting a middleware machine environment.

Background:

[0002] Infiniband (IB) Architecture is a communications and management infrastructure that supports both I/O and interprocessor communications for one or more computer systems. An IB Architecture system can scale from a small server with a few processors and a few I/O devices to a massively parallel installation with hundreds of processors and thousands of I/O devices.

[0003] The IB Architecture defines a switched communications fabric allowing many devices to concurrently communicate with high bandwidth and low latency in a protected, remotely managed environment. An end node can communicate with over multiple IB Architecture ports and can utilize multiple paths through the IB Architecture fabric. A multiplicity of IB Architecture ports and paths through the network are provided for both fault tolerance and increased data transfer bandwidth.

[0004] These are the generally areas that embodiments of the invention are intended to address.

Summary:

[0005] Described herein is a system and method for performing subnet initialization in a middleware machine environment. In accordance with one embodiment, a plurality of management nodes in the middleware machine environment can be interconnected so that a plurality of subnet managers that are associated with the plurality of management nodes can communicate with each other using an in-band communication protocol. The plurality of subnet managers can negotiate and elect a master subnet manager, which operates to configure and manage the middleware machine environment. The master subnet manager can replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol. The master subnet manager then sets up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

[0006] In one aspect, provided is a subnet manager in a middleware machine environment, comprising: a communication module configured to communicating with other subnet managers using an in-band communication protocol; an electing module configured to negotiate with the other subnet managers and elect a master subnet manager; a replicating module configured to, responsive to the subnet magener being elected as the master subnet manager, replicate one or more subnet initialization policies to the other subnet managers in the middleware machine environment using the in-band communication protocol; and a setting up module configured to, responsive to the subnet magener being elected as the master subnet manager, set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

[0007] In some embodiments, the in-band communication protocol provides communication redundancy in addition to Ethernet networking in the middleware machine environment.

[0008] In some embodiments, one or more partition policies can define a partition configuration in the middleware machine environment, and wherein the one or more partition policies can be supplied to the middleware machine environment through an initialization policy transaction.

[0009] In some embodiments, the subnet manager further comprises an initialization module configured to, responsive to the subnet manager being elected as the master subnet manager, use a default patitioning policy for initialization when no partitioning policy is specified.

[00010] In some embodiments, the subnet manager further comprises a synchronization module configured to, responsive to the subnet manager being elected as the master subnet manager, synchronize fabric level configuration policy by starting up initially using a default partition policy, and proceeding to install a partition configuration in the middleware machine environment, when fabric level synchronization has been achieved.

[00011] In another aspect, provided is a management node in a middleware machine environment, comprising the above-mentioned subnet manager .

[00012] In still another aspect, provided is a system for performing subnet initialization in a middleware machine environment, comprising a plurality of the above-mentioned mangement nodes which are interconnected in the middleware machine environment.

[00013] In some embodiments, the plurality of management nodes are part of an infiniband (IB) subnet.

[00014] In some embodiments, the plurality of management nodes include one or more network switches, wherein each said subnet manager resides on a network switch. [00015] In some embodiments, the system further comprises a seperate storage system that connects with the plurality of management nodes.

[00016] In some embodiments, the system further comprises a plurality of host servers that connect with the plurality of management nodes.

[00017] In some embodiments, the system further comprises one or more gateway instances that can be accessed by a guest.

[00018] In some embodiments, the system further comprises a plurality of policy daemons, wherein each said policy daemon is associated with a subnet manager.

[00019] In some embodiments, the policy daemon that collocates with the master subnet manager is responsible for configuring and managing the middleware machine environment using one or more policies.

[00020] In some embodiments, the master subnet manager operates to leave current partition membership unchanged for nodes that are already operational and only ensures that a predefined partition is set up for new management nodes.

[00021] In some embodiments, a subnet manager operates to fully initialize all end-nodes with a complete partition membership as defined in a partitioning policy, when the partition policy has been fully synchronized for said management nodes.

[00022] In yet still another aspect, provided is a system for performing subnet initialization in a middleware machine environment, comprising: a plurality of management nodes interconnected in the middleware machine environment, a plurality of subnet managers that are associated with the plurality of management nodes, wherein the plurality of management nodes allows the plurality of subnet managers to communicate with each other using an in-band communication protocol, wherein the plurality of subnet managers negotiates with each other and elect a subnet manager as a master subnet manager, and wherein the master subnet manager operates to replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol, and set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies. Brief Description of the Figure:

[00023] Figure 1 shows an illustration of an exemplary configuration for a middleware machine, in accordance with an embodiment of the invention.

[00024] Figure 2 shows an illustration of a middleware machine environment, in accordance with an embodiment of the invention.

[00025] Figure 3 shows an illustration of a system for performing subnet initialization in a middleware machine environment, in accordance with an embodiment of the invention.

[00026] Figure 4 illustrates an exemplary flow chart for performing subnet initialization in a middleware machine environment in accordance with an embodiment.

[00027] Figure 5 shows an illustration of a middleware machine environment with different partition groups, in accordance with an embodiment of the invention.

[00028] Figure 6 shows an explanatory functional block diagram of a subnet manager in a middleware machine environment, in accordance with some embodiments of the invention as described above.

[00029] Figure 7 shows an explanatory functional block diagram of a system for performing subnet initialization in a middleware machine environment, in accordance with some embodiments of the invention.

Detailed Description:

[00030] Described herein is a system and method for providing a middleware machine or similar platform. In accordance with an embodiment of the invention, the system comprises a combination of high performance hardware (e.g. 64-bit processor technology, high performance large memory, and redundant InfiniBand and Ethernet networking) together with an application server or middleware environment, such as WebLogic Suite, to provide a complete Java EE application server complex which includes a massively parallel in-memory grid, that can be provisioned quickly, and that can scale on demand. In accordance with an embodiment of the invention, the system can be deployed as a full, half, or quarter rack, or other configuration, that provides an application server grid, storage area network, and InfiniBand (IB) network. The middleware machine software can provide application server, middleware and other functionality such as, for example, WebLogic Server, JRockit or Hotspot JVM, Oracle Linux or Solaris, and Oracle VM. In accordance with an embodiment of the invention, the system can include a plurality of compute nodes, one or more IB switch gateways, and storage nodes or units, communicating with one another via an IB network. When implemented as a rack configuration, unused portions of the rack can be left empty or occupied by fillers.

[00031] In accordance with an embodiment of the invention, referred to herein as "Sun Oracle Exalogic" or "Exalogic", the system is an easy-to-deploy solution for hosting middleware or application server software, such as the Oracle Middleware SW suite, or Weblogic. As described herein, in accordance with an embodiment the system is a "grid in a box" that comprises one or more servers, storage units, an IB fabric for storage networking, and all the other components required to host a middleware application. Significant performance can be delivered for all types of middleware applications by leveraging a massively parallel grid architecture using, e.g. Real Application Clusters and Exalogic Open storage. The system delivers improved performance with linear I/O scalability, is simple to use and manage, and delivers mission-critical availability and reliability. [00032] Figure 1 shows an illustration of an exemplary configuration for a middleware machine, in accordance with an embodiment of the invention. As shown in Figure 1 , the middleware machine 100 uses a single rack configuration that includes two gateway network switches, or leaf network switches, 102 and 103 that connect to twenty-eight server nodes. Additionally, there can be different configurations for the middleware machine. For example, there can be a half rack configuration that contains a portion of the server nodes, and there can also be a multi-rack configuration that contains a large number of servers.

[00033] As shown in Figure 1 , the server nodes can connect to the ports provided by the gateway network switches. As shown in Figure 1 , each server machine can have connections to the two gateway network switches 102 and 103 separately. For example, the gateway network switch 102 connects to the port 1 of the servers 1-14 106 and the port 2 of the servers 15-28 107, and the gateway network switch 103 connects to the port 2 of the servers 1-14 108 and the port 1 of the servers 15-28 109.

[00034] In accordance with an embodiment of the invention, each gateway network switch can have multiple internal ports that are used to connect with different servers, and the gateway network switch can also have external ports that are used to connect with an external network, such as an existing data center service network.

[00035] In accordance with an embodiment of the invention, the middleware machine can include a separate storage system 1 10 that connects to the servers through the gateway network switches. Additionally, the middleware machine can include a spine network switch 101 that connects to the two gateway network switches 102 and 103. As shown in Figure 1 , there can be optionally two links from the storage system to the spine network switch.

IB Fabric/Subnet

[00036] In accordance with an embodiment of the invention, an IB Fabric/Subnet in a middleware machine environment can contain a large number of physical hosts or servers, switch instances and gateway instances that are interconnected in a fat-tree topology.

[00037] Figure 2 shows an illustration of a middleware machine environment, in accordance with an embodiment of the invention. As shown in Figure 2, the middleware machine environment 200 includes an IB subnet or fabric 220 that connects with a plurality of end nodes. The IB subnet includes a plurality of subnet managers 211-214, each of which resides on one of a plurality of network switches 201-204. The subnet managers can communicate with each other using an in-band communication protocol 210, such as the Management Datagram (MAD) / Subnet Management Packet (SMP) based protocols or other protocol such as the Internet Protocol over IB (IPolB).

[00038] In accordance with an embodiment of the invention, a single IP subnet can be constructed on the IB fabric allowing the switches to communicate securely among each other in the same IB fabric (i.e. full connectivity among all switches). The fabric based IP subnet can provide connectivity between any pair of switches when at least one route with operational links exists between the two switches. Recovery from link failures can be achieved if an alternative route exists by re-routing.

[00039] The management Ethernet interfaces of the switches can be connected to a single network providing IP level connectivity between all the switches. Each switch can be identified by two main IP addresses: one for the external management Ethernet and one for the fabric based IP subnet. Each switch can monitor connectivity to all other switches using both IP addresses, and can use either operational address for communication. Additionally, each switch can have a point-to-point IP link to each directly connected switch on the fabric. Hence, there can be at least one additional IP address.

[00040] IP routing setups allow a network switch to route traffic to another switch via an intermediate switch using a combination of the fabric IP subnet, the external management Ethernet network, and one or more fabric level point-to-point IP links between pairs of switches. IP routing allows external management access to a network switch to be routed via an external Ethernet port on the network switch, as well as through a dedicated routing service on the fabric.

[00041] The IB fabric includes multiple network switches with managment Ethernet access to a managment network. There is in-band physical connectivity between the switches in the fabric. In one example, there is at least one in-band route of one or more hops between each pair of switches, when the IB fabric is not degraded. Management nodes for the IB fabric include network switches and management hosts that are connected to the IB fabric.

[00042] A subnet manager can be accessed via any of its private IP addresses. The subnet manager can also be accessible via a floating IP address that is configured for the master subnet manager when the subnet manager takes on the role as a master subnet manager, and the subnet manager is un-configured when it is explicitly released from the role. A master IP address can be defined for both the external management network as well as for the fabric based management IP network. No special master IP address needs to be defined for point-to-point IP links.

[00043] In accordance with an embodiment of the invention, each physical host can be virtualized using virtual machine based guests. There can be multiple guests existing concurrently per physical host, for example one guest per CPU core. Additionally, each physical host can have at least one dual-ported Host Channel Adapter (HCA), which can be virtualized and shared among guests, so that the fabric view of a virtualized HCA is a single dual-ported HCA just like a non-virtualized/shared HCA.

[00044] The IB fabric can be divided into a dynamic set of resource domains implemented by IB partitions. Each physical host and each gateway instance in an IB fabric can be a member of multiple partitions. Also, multiple guests on the same or different physical hosts can be members of the same or different partitions. The number of the IB partitions for an IB fabric may be limited by the P_Key table size.

[00045] In accordance with an embodiment of the invention, a guest may open a set of virtual network interface cards (vNICs) on two or more gateway instances that are accessed directly from a vNIC driver in the guest. The guest can migrate between physical hosts while either retaining or having updated vNIC associates.

[00046] In accordance with an embodiment of the invention, switchs can start up in any order and can dynamically select a master subnet manager according to different negotiation protocols, for example an IB specified negotiation protocol. If no partitioning policy is specified, a default partitioning enabled policy can be used. Additionally, the management node partition and the fabric based management IP subnet can be established independently of any additional policy infomation and independently of whether the complete fabric policy is known by the master subnet manager. In order to allow fabric level configuration policy information to be synchronized using the fabric based IP subnet, the subnet manager can start up initially using the default partition policy. When fabric level synchronization has been achived, the partition configuration, which is current for the fabric, can be installed by the master subnet manager. Partial subnet initialization

[00047] In accordance with an embodiment of the invention, subnet initialization can be performed in a middleware machine environment. A plurality of management nodes in the middleware machine environment can be interconnected so that a plurality of subnet managers that are associated with the plurality of management nodes can communicate with each other using an in-band communication protocol. The plurality of subnet managers can negotiate and elect a master subnet manager, which operates to configure and manage the middleware machine environment. The master subnet manager can replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol. The master subnet manager can then set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

[00048] Figure 3 shows an illustration of a system for performing subnet initialization in a middleware machine environment, in accordance with an embodiment of the invention. As shown in Figure 3, the middleware machine environment 300 includes an IB subnet or fabric 320 that manages a plurality of client nodes or end nodes. The IB subnet includes a plurality of management nodes 301-304 that maintains a plurality of subnet managers 321-324 associated with a plurality of configuration daemon 311 -314. The management nodes can be either a network switch or a predefined host server. The subnet managers can communicate with each other using an in-band communication protocol 310, such as the Internet Protocol over Infiniband (IPolB).

[00049] In accordance with an embodiment of the invention, the system can provide a plurality of policy daemons 311-314, each of which is associated with a subnet manager 321 -324. In the example as shown in Figure 3, the policy daemon 311 that collocates with the master subnet manager 301 is responsible for configuring and managing the end nodes in the middleware machine environment using one or more policies.

[00050] One exemplary policy managed by a policy daemon in a middleware machine environment can be a partition configuration policy. For example, a middleware machine environment that includes end nodes, A, B and C can be partitioned into two groups: a Group I that includes nodes A and B and a Group II that includes node C. A partition configuration policy can define a partition update that requires deleting node B from the Group I, before adding node B into the Group II. This partition configuration policy can require that the master subnet manager will not allow a new partition to add node B into Group II without first deleting nodes B from Group I. This partition configuration policy can be enforced by the master subnet manager using a policy daemon.

[00051] The in-band communication connectivity provides communication redundancy in addition to the Ethernet networking in the middleware machine environment. The system can take advantage of the communication redundancy provided by the middleware machine environment and use the IB subnet under configuration as the management network for distributing fabric initialization policies in order to fully configure the middleware machine environment.

[00052] In accordance with an embodiment of the invention, the subnet managers in the IB subnet can first negotiate with each other and elect a master subnet manager 321 , which is responsible for configuring and managing the middleware machine environment. The master subnet manager can replicate one or more fabric initialization policies in the configuration daemon to other subnet manager master candidates 322-324 in the middleware machine environment, using the in-band communication protocol within the context of an initialization policy transaction 309. Then, the master subnet manager can set up the plurality of end nodes in the middleware machine environment based on the one or more subnet initialization policies.

[00053] In accordance with an embodiment of the invention, each management node can connect with one or more client nodes or end nodes such as the host servers within the middleware machine environment, using Ethernet network connection 330. The network switch can be either a leaf switch that communicates directly with the end nodes, or a spine switch that communicates with the end nodes through the leaf switches. The network switches can communicate with the host servers via the switch ports of the network switches and host ports of the host servers. In IB network, partitions can be defined to specify which end ports are able to communicate with other end ports. The middleware machine environment can employ a fat-tree topology, which allows for a small number of switches sitting at the top layers of the fat tree while maintaining a large number of end nodes as leafs of the tree.

[00054] Figure 4 illustrates an exemplary flow chart for performing subnet initialization in a middleware machine environment in accordance with an embodiment. As shown in Figure 4, at step 401 , a plurality of management nodes can be interconnected in the middleware machine environment so that a plurality of subnet managers associated with the management nodes can communicate with each other using an in-band communication protocol. The plurality of subnet managers can first negotiate among each other and elect a master subnet manager, at step 402. The master subnet manager is responsible for configuring and managing the middleware machine environment. Furthermore, at step 403, the master subnet manager can replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol. Then, at step 404, the master subnet manager can set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

Subnet initialization with partition configuration policies

[00055] Figure 5 shows an illustration of a middleware machine environment with different partition groups, in accordance with an embodiment of the invention. As shown in Figure 5, one or more partition policies 519 can define a partition configuration in a middleware machine environment 500. The one or more partition policies can be supplied to the middleware machine environment through an initialization policy transaction 509. As a result of the partition configuration, the middleware machine environment can be divided into three separate partition groups: Partition A 531 , Partition B 532, and Partition C 533. Each partition group can be associated with a unique partition key and can be logically connect to the IB subnet 520 using a separate connection 541-543.

[00056] In accordance with an embodiment of the invention, the subnet managers located at the network switches in an IB network can communicate with the host servers via the switch ports of the network switches and host ports of the host servers. Partitions can be defined to specify which end ports are able to communicate with each other. An IB partition can be defined by a P_Key which is a sixteen bit numerical value, where the lower fifteen bits define the partition number, and the top bit defines whether the associated port is a full or limited member in that partition. A limited member can only communicate with a full member, whereas full members can communicate among themselves as well as with any limited member of the same partition. [00057] Each end port can be a member of at least one partition in order to be able to communicate with any other end-port (i.e. use of the P_Key field is not optional in the IB transport protocol). For switch ports, partition enforcement is an optional feature controlled by the Subnet Manager.

[00058] The subnet manager controls partition membership by updating the partition tables of the relevant end ports and switch ports. If the trust model for the fabric or a particular host implies that the host or host administrator may not be fully adhering to the partitioning that the subnet manager has defined, then switch-based enforcement can be used so that the switch ingress/egress port can enforce the same restrictions as the subnet manager has set up for the corresponding host port.

[00059] The number of partitions that a port can represent is a hardware implementation parameter reflected as port attributes to the subnet manager (For example, in the 16-64/128 range). Hence, while it is possible to have up to 32K individual P_Key values defined for the fabric, each port can be a member of a limited subset, and the switch enforcement of partitioning is enabled for switch fabric ingress/egress ports that are directly connecting to end-ports.

[00060] The default partition key (OxFFFF - full, or 0x7FFF -limited) can be reserved by the specification and is used (by default) for access to the Subnet Manager. In a system that uses partition enforcement, end nodes are limited members of the default partition in order to be able to communicate with the subnet manager. If partition enforcement is not used, then end-nodes are full members of the default partition. The value zero can be reserved and used to represent an invalid P_Key value.

[00061] In order to synchronize fabric level configuration policy information among the fabric based subnet, the master subnet manager can start up initially using the default partition policy. When fabric level synchronization has been achieved, the master subnet manager can proceed to install the partition configuration in the middleware machine environment.

[00062] In accordance with an embodiment of the invention, using the default configuration with end nodes only being limited members of the default partition allows no illegal communication between end-nodes before the correct current partitioning policy can be implemented. The master subnet manager can set up the management nodes to be full members of a predefined partition and thereby allow the policy daemons to communicate using IPolB, so that replication/synchronization of the current dynamic partition policy can take place via the IB fabric. However, such a scheme forcing end nodes to become only limited members of the default partition has the disadvantage of interrupting existing communication, when the fabric is already operational with on-going communication between end nodes at the time the current master subnet manager takes over the subnet.

[00063] In accordance with an embodiment of the invention, in order to avoid an interruption of on-going legal communication, and to provide IPolB based communication between all designated management nodes before the policy daemons have been able to synchronize, the new subnet manager can leave current partition membership for already operational nodes unchanged and can ensure that only the predefined partition is set up for the management nodes. Non-management nodes that are currently not operational may not be initialized (i.e. requires initialization from the master subnet manager). Or, alternatively, the non-management nodes may only be set up with limited membership in the default partition. Then, when the current partition policy has been fully synchronized, all end-nodes can be fully initialized with the complete partition membership defined by the current partitioning policy.

[00064] The above scheme can be useful since any current master subnet manager can initialize the IB fabric to be in compliance with the current partition policy, and the failure of a master subnet manager can leave the fabric in a state where it is either fully initialized or is in a state where the current policy have been partially implemented. This includes the case where an old master fails before it has completed removing membership defined by an old policy but not present in a new policy.

[00065] In accordance with an embodiment of the invention, the replication and synchronization between the policy daemons can ensure that a quorum or majority of the subnet manager master candidates are updated with the new policy before any change is committed. Then, the actual removing of old partition membership and addition of new membership can be completely implemented by a single current master subnet manager, or be implemented in two or more partial steps by two or more master subnet manager instances as the result of fail-over or hand-over of mastership in the subnet. Each such partial step may or may not change any state in the IB subnet but can be guaranteed to not introduce any intermediate states that are not either a subset of the membership defined by the old policy or a subset of the new policy without any chance of mixing any policy from the two versions. Also, in the case of one or more fail-over or hand-over events, the new master subnet manager may not undo the operations performed by the former master but can start from the relevant next step based on the current policy and the current state of the IB subnet that the new master discovers dynamically.

[00066] In accordance with some embodiments, Figure 6 shows an explanatory functional block diagram of a subnet manager 600 configured in accordance with the principles of the invention as described above, in a middleware machine environment, and Figure 7 shows an explanatory functional block diagram of a subnet initialization system 700 configured in accordance with the principles of the invention as described above, comprising a plurality of management nodes 702 each of which comprises a subnet manger 600 as shown in Figure 6. The functional blocks of the subnet manager and subnet initialization system can be implemented by hardware, software, or a combination of hardware and software to carry out the principles of the invention. It is understood by persons of skill in the art that the functional blocks described in Figures 6 and 7 may be combined or separated into sub-blocks to implement the principles of the invention. Therefore, the description herein may support any possible combination or separation or further definition of the functional blocks described herein.

[00067] As shown in Figure 6, the subnet manager 600 can comprise a communication module 602, an electing module 604, a replicating module 606 and a setting up module 608. The communication module 603 can be configured to communicating with other subnet managers using an in-band communication protocol. The electing module 604 can be configured to negotiate with the other subnet managers and elect a master subnet manager. The replicating module 606 can be configured to, responsive to the subnet magener 600 being elected as the master subnet manager, replicate one or more subnet initialization policies to the other subnet managers in the middleware machine environment using the in-band communication protocol. The setting up module 608 can be configured to, responsive to the subnet magener 600 being elected as the master subnet manager, set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

[00068] In some embodiments, the in-band communication protocol can provide communication redundancy in addition to Ethernet networking in the middleware machine environment.

[00069] In some embodiments, one or more partition policies can define a partition configuration in the middleware machine environment, and the one or more partition policies can be supplied to the middleware machine environment through an initialization policy transaction, such as the initialization policy transaction 509 in Figure 5.

[00070] In some embodiments, the subnet manager 600 can optionally comprise an initialization module 610. The initialization module 610 can be configured to, responsive to the subnet manager 600 being elected as the master subnet manager, use a default patitioning policy for initialization when no partitioning policy is specified.

[00071] In some embodiments, the subnet manager 600 can optionally comprise a synchronization module 610. The synchronization module 610 can be configured to, responsive to the subnet manager 600 being elected as the master subnet manager, synchronize fabric level configuration policy by starting up initially using a default partition policy, and proceeding to install a partition configuration in the middleware machine environment, when fabric level synchronization has been achieved.

[00072] As shown in Figure 7, a subnet initialization system 700 can comprise a plurality of management nodes 702. Each management node 702 can comprise a subnet manager 600 as shown in Figure 6 and the management nodes 702 can be interconnected in the middleware machine environment. It should be noted that, although the subnet manager 600 is shown in Figure 7 as included in the management node 702, it can be otherwise associated with the management node 702. For example, the subnet manager 600 can be physically separated from the management node 702 and communicate with other subnet managers through the associated management node. The plurality of management nodes 702 can be part of an infiniband (IB) subnet.

[00073] In some embodiments, although not shown in Figure 7, the plurality of management nodes 702 can include one or more network switches, wherein each subnet manager resides on a network switch.

[00074] In some embodiments, the subnet initialization system 700 can optionally comprise a seperate storage system 704. Although not shown, the storage system 704 can connect with the management nodes 702.

[00075] In some embodiments, the subnet initialization system 700 can optionally comprise a plurality of host servers 706. Although not shown, the host servers 706 can connect with the management nodes 702.

[00076] In some embodiments, the subnet initialization system 700 can optionally comprise one or more gateway instances 708 that can be accessed by a guest. And in some embodiments, the subnet initialization system 700 can optionally comprise a plurality of policy daemons 710, wherein each said policy daemon 710 is associated with a subnet manager 600. The policy daemon 710 that collocates with the master subnet manager can be responsible for configuring and managing the middleware machine environment using one or more policies. It should be noted that, although the gateway instance 708 and the policy daemon 710 are shown as separate components from the management node 702 in Figure 7, they can also be arranged in or reside on the management node 702.

[00077] In some embodiments, the master subnet manager in the subnet initialization system 700 can operate to leave current partition membership unchanged for nodes that are already operational and only ensures that a predefined partition is set up for new management nodes. Alternatively, this operation can be performed by a special module of the master subnet manager or by one of the modules 602-612 as described with reference to Figure 6.

[00078] In some embodiments, a subnet manager 600 in the subnet initialization system 700 can operate to fully initialize all end-nodes with a complete partition membership as defined in a partitioning policy, when the partition policy has been fully synchronized for the management nodes 702. Alternatively, this operation can be performed by a special module of the subnet manager 600 or by one of the modules 602-612 as described with reference to Figure 6.

[00079] The present invention may be conveniently implemented using one or more conventional general purpose or specialized digital computer, computing device, machine, or microprocessor, including one or more processors, memory and/or computer readable storage media programmed according to the teachings of the present disclosure. Appropriate software coding can readily be prepared by skilled programmers based on the teachings of the present disclosure, as will be apparent to those skilled in the software art.

[00080] In some embodiments, the present invention includes a computer program product which is a storage medium or computer readable medium (media) having instructions stored thereon/in which can be used to program a computer to perform any of the processes of the present invention. The storage medium can include, but is not limited to, any type of disk including floppy disks, optical discs, DVD, CD-ROMs, microdrive, and magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic or optical cards, nanosystems (including molecular memory ICs), or any type of media or device suitable for storing instructions and/or data.

[00081] The foregoing description of the present invention has been provided for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations will be apparent to the practitioner skilled in the art. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications that are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the following claims and their equivalence.

Claims

Claims:

What is claimed is: 1. A system for performing subnet initialization in a middleware machine environment, comprising:

one or more microprocessors;

a plurality of management nodes interconnected in the middleware machine environment that run on the one or more microprocessors,

a plurality of subnet managers that are associated with the plurality of management nodes, wherein the plurality of management nodes allows the plurality of subnet managers to communicate with each other using an in-band communication protocol, wherein the plurality of subnet managers negotiate with each other and elect a subnet manager as a master subnet manager, and

wherein the master subnet manager operates to:

replicate one or more subnet initialization policies to other subnet managers in the middleware machine environment using the in-band communication protocol, and set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

2. The system according to Claim 1 , wherein:

the plurality of management nodes are part of an Infiniband (IB) subnet.

3. The system according to Claim 1 or 2, wherein:

the plurality of management nodes include one or more network switches, wherein each said subnet manager resides on a network switch.

4. The system according to any one of Claims 1 to 3, further comprising:

a separate storage system that connects with the plurality of management nodes.

5. The system according to any one of Claims 1 to 4, further comprising:

a plurality of host servers that connect with the plurality of management nodes.

6. The system according to any one of Claims 1 to 5, further comprising:

one or more gateway instances that can be accessed by a guest.

7. The system according to any one of Claims 1 to 6, further comprising: a plurality of policy daemons, wherein each said policy daemon is associated with a subnet manager.

8. The system according to Claim 7, wherein:

the policy daemon that collocates with the master subnet manager is responsible for configuring and managing the middleware machine environment using one or more policies.

9. The system according to any one of Claims 1 to 8, wherein:

the in-band communication protocol provides communication redundancy in addition to

Ethernet networking in the middleware machine environment.

10. The system according to any one of Claims 1 to 9, wherein:

one or more partition policies can define a partition configuration in the middleware machine environment, and wherein the one or more partition policies can be supplied to the middleware machine environment through an initialization policy transaction.

1 1. The system according to any one of Claims 1 to 10, wherein:

the master subnet manager can use a default partitioning policy for initialization when no partitioning policy is specified.

12. The system according to any one of Claims 1 to 1 1 , wherein:

the master subnet manager operates to synchronize fabric level configuration policy by starting up initially using a default partition policy, and proceeding to install a partition configuration in the middleware machine environment, when fabric level synchronization has been achieved.

13. The system according to any one of Claims 1 to 12, wherein:

the master subnet manager operates to leave current partition membership unchanged for nodes that are already operational, wherein the master subnet manager ensures that a predefined partition is set up for new management node allowing all management nodes to communicate and synchronize via a communication protocol.

14. The system according to any one of Claims 1 to 13, wherein:

a subnet manager operates to fully initialize all end-nodes with a complete partition membership as defined in a partitioning policy, when the partition policy has been fully synchronized for said management nodes.

15. A method for performing subnet initialization in a middleware machine environment, comprising:

interconnecting a plurality of management nodes running on one or more microprocessors in the middleware machine environment, wherein the plurality of management nodes allows a plurality of subnet managers that are associated with the plurality of management nodes to communicate with each other using an in-band communication protocol;

negotiating among the plurality of subnet managers and electing a subnet manager as a master subnet manager, wherein the master subnet manager operates to configure and manage the middleware machine environment;

replicating one or more subnet initialization policies, via the master subnet manager, to other subnet managers in the middleware machine environment using the in-band communication protocol; and

setting up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

16. The method according to Claim 15, wherein:

the plurality of management nodes are part of an Infiniband (IB) subnet.

17. The method according to Claim 15 or 16, wherein:

18. The method according to any of Claims 15 to 17, further comprising:

providing a seperate storage system that connects with the plurality of management nodes.

19. The method according to any of Claims 15 to 18, further comprising:

20. The method according to any of Claims 15 to 19, further comprising:

providing one or more gateway instances that can be accessed by a guest.

21. The method according to any of Claims 15 to 20, further comprising:

providing a plurality of policy daemons, wherein each said policy daemon is associated with a subnet manager.

22. The method according to Claim 21 , wherein:

23. The method according to any of Claims 15 to 22, further comprising:

the in-band communication protocol providing communication redundancy in addition to Ethernet networking in the middleware machine environment.

24. The method according to any of Claims 15 to 23, further comprising:

providing one or more partition policies that define a partition configuration in the middleware machine environment, and wherein the one or more partition policies can be supplied to the middleware machine environment through an initialization policy transaction.

25. The method according to any of Claims 15 to 24, further comprising:

the master subnet manager using a default patitioning policy for initialization when no partitioning policy is specified.

26. The method according to any of Claims 15 to 25, further comprising:

the master subnet manager operating to synchronize fabric level configuration policy by starting up initially using a default partition policy, and proceeding to install a partition configuration in the middleware machine environment, when fabric level synchronization has been achieved.

27. The method according to any of Claims 15 to 26, further comprising:

the master subnet manager operating to leave current partition membership unchanged for nodes that are already operational and only ensuring that a predefined partition is set up for new management nodes.

28. The method according to any preceding Claim, further comprising:

a subnet manager operating to fully initialize all end-nodes with a complete partition membership as defined in a partitioning policy, when the partition policy has been fully synchronized for said management nodes.

29. A computer program comprising program readable instructions that when loaded into and executed by one or more computer systems cause the one or more computer systems to perform the method of any of claims 15 to 28.

30. A computer program product comprising a computer readable storage medium storing the computer program of claim 29.

31. A machine readable medium having instructions stored thereon that when executed cause a system to perform the steps of:

32. A computer program for causing a computer to perform the steps of:

33. A subnet manager in a middleware machine environment, comprising:

a communication module configured to communicating with other subnet managers using an in-band communication protocol; an electing module configured to negotiate with the other subnet managers and elect a master subnet manager;

a replicating module configured to, responsive to the subnet magener being elected as the master subnet manager, replicate one or more subnet initialization policies to the other subnet managers in the middleware machine environment using the in-band communication protocol; and

a setting up module configured to, responsive to the subnet magener being elected as the master subnet manager, set up a plurality of client nodes in the middleware machine environment based on the one or more subnet initialization policies.

34. The subnet manager according to Claim 33, wherein:

the in-band communication protocol provides communication redundancy in addition to Ethernet networking in the middleware machine environment.

35. The subnet manager according to Claim 33, wherein:

36. The subnet manager according to Claim 33, further comprising:

an initialization module configured to, responsive to the subnet manager being elected as the master subnet manager, use a default patitioning policy for initialization when no partitioning policy is specified.

37. The subnet manager according to Claim 33, further comprising:

a synchronization module configured to, responsive to the subnet manager being elected as the master subnet manager, synchronize fabric level configuration policy by starting up initially using a default partition policy, and proceeding to install a partition configuration in the middleware machine environment, when fabric level synchronization has been achieved.

38. A management node in a middleware machine environment, comprising a subnet manager according to any of Claims 33-37.

39. A system for performing subnet initialization in a middleware machine environment, comprising a plurality of mangement nodes according to Claim 38 which are interconnected in the middleware machine environment.

40. The system according to claim 39, wherein:

the plurality of management nodes are part of an Infiniband (IB) subnet.

41. The system according to Claim 39, wherein:

42. The system according to Claim 39, further comprising:

a seperate storage system that connects with the plurality of management nodes.

43. The system according to Claim 39, further comprising:

44. The system according to Claim 39, further comprising:

one or more gateway instances that can be accessed by a guest.

45. The system according to Claim 39, further comprising:

a plurality of policy daemons, wherein each said policy daemon is associated with a subnet manager.

46. The system according to Claim 45, wherein:

47. The system according to claim 39, wherein:

the master subnet manager operates to leave current partition membership unchanged for nodes that are already operational and only ensures that a predefined partition is set up for new management nodes.

48. The system according to claim 39, wherein:

49. A system for performing subnet initialization in a middleware machine environment, comprising:

a plurality of management nodes interconnected in the middleware machine environment,

a plurality of subnet managers that are associated with the plurality of management nodes, wherein the plurality of management nodes allows the plurality of subnet managers to communicate with each other using an in-band communication protocol, wherein the plurality of subnet managers negotiates with each other and elect a subnet manager as a master subnet manager, and

wherein the master subnet manager operates to