Canisters for Node

Each processing unit is one node canister that is known as the term "node. The two nodes in the control enclosure form an I/O cluster that is linked to SAN fabric. Each processing unit is an IBM node-canister that is also known as the term "node. Two nodes in the enclosure create an I/O cluster that connects host systems and, occasionally, external storage systems.

In order to ensure redundancy, the nodes are arranged in pairs, which form an entire system. The system could have one or four nodes. Every pair is referred to by the name of an I/O Group. Each node is able to be part of the same I/O group. There can be a maximum of four I/O groups, each with two nodes are supported.At the same moment, a single node within the system is responsible for the configuration activities. This node handles the storage of details of the configuration, which defines the system's configuration and acts as a hub for commands to configure. If the node that is configured fails, a different node within the system takes the responsibility.
Naming objects

All objects within the system are named by names that are defined by the user or generated by the system.

Make sure you choose a meaningful name before creating an object. If you don't choose an appropriate name for an object it will generate the name for you. A name that is well-chosen serves not just as a name for the objectbut can also serve serves as an instrument to monitor and control the object. The choice of a meaningful name is essential if you opt to utilize the configuration backup and restore.
Naming rules

The following guidelines apply when choosing the name you want to give an object:

Don't begin names using underscores even when it's possible. The use of underscores as the first character in an inscription is a reserved convention used by the process of restoring configurations to the system.

The first letter of the file volume name can't be a numeric number.

The name may be up to 63 characters, with a single exception. The lsfabric command shows long names for objects that are reduced by 15 characters to accommodate the nodes and systems.

Acceptable characters include uppercase letters (A Z) and lowercase letters (a Z - A) and numbers (0 9) underscore (_) as well as the period (. ) and hyphen (-), and spaces.

Names cannot start or end with the letter "A" or.

Names of objects have to be distinct within the type of object. For instance volumes and MDisk could each be referred to as ABC but two volumes can't be called ABC.

The default name of the object is valid (object prefixes with the number).

The objects can be renamed according to the current name.
Overview of the object

Before setting the system up, it is essential to be aware of the concepts and objects that are in the surrounding.

Each processing unit functions as a canister for nodes, which is also referred to as an individual node. Two nodes in the control enclosure create an I/O group connected with the SAN fabric. Two nodes inside the enclosure create an I/O cluster that connects host systems, and additionally, other storage devices.

The volumes are disks that contain logical information which are displayed to the system. Both canisters for nodes give an access point to volumes. When an application server executes I/O to a volume it will be able to connect to the volume through either of the nodes within the group called I/O.

The canisters of node are connected to drives in enclosures. These drives make at least one Redundant arrays of independent disks (RAID). The arrays can also be referred to by the name managed disks (MDisks).

The system also has the ability to recognize the logical units which are displayed by storage systems that are used for back-end storage as MDisks.

Each MDisk is split into several extents.

MDisks are organized into groups. These groups are referred to as storage pools.

The system also recognizes the logical units which are displayed through Fibre Channel external storage systems as MDisks.

Each volume is comprised of two or one volume copies. Every volume is an individual physical replica of the data that is kept within the volume. A volume that has two copies is referred to as the mirror volume. Volume copies are composed of MDisk sizes. Each MDisk that contributes to a specific volume copy must be part of the same storage pool.

A volume could be provisioned thinly.By the capacity. Volumes with thin provisioning may be created to automatically expand their capacity by allocating more space.

At any given moment, just one of the nodes in the system is able to handle the configuration tasks. The node responsible for managing configuration activities is known as"the configuration" node. It's accountable for maintaining an inventory of information that defines the system's configuration. It is also an important place to set up.

In the event of an SCSI using the Fibre Channel and an Fibre Channel over Ethernet (FCoE) connection, the canisters for nodes be able to recognize FC and FCoE ports that are connected to SAN.You can build virtual hosts that combine WWPNs that are associated with one server for applications, or one of them.

Host servers are able to access volumes using Fibre Channel, iSCSI, or SAS. The system uses World-wide ports (WWPNs) to determine the ports of hosts servers. iSCSI, SAS, and Fibre Channel ports are available on hosts servers. The host server makes use of its iSCSI accredited name (IQN) to identify iSCSI hosts.

To create the SCSI through an Ethernet connection to establish the SCSI via Ethernet connection. The iSCSI qualified number (IQN) signifies the iSCSI adapter is the target (destination) adapter. Hosts can be fitted with either IQNs together with WWPNs.

System hosts are virtualized versions of host systems that are physical and applications servers are allowed access to the system volume. Every definition for a host system specifies the method of connection and the port, or IQN and also the volume which host applications can access.

System (IQN) hosts are virtual hosts with physical hosts that share the same volume. Host objects are created which are WWPNs grouped together or IQNs that are part of one application server or to a set of them.

The system can provide block-level aggregate as well as management of the disks that are stored within the SAN. The system can manage a range of storage systems that act as back-end systems and converts physical data within those storage units into digital disks which are accessible by desktop computers as well as application servers in the SAN. The SAN is designed to ensure that application servers cannot access the physical storage on the back end. This prevents conflict with the system , or application servers who are trying to control the storage on the back end.

Post a Comment

Previous Post Next Post