Optimizing operations by enabling collocation of client files

Collocation of client files reduces the number of volume mounts that are required when users restore, retrieve, or recall many files from a storage pool. Collocation thus reduces the amount of time that is required for these operations.

About this task

With collocation enabled, the server tries to keep files on a minimal number of sequential-access storage volumes. The files can belong to a single client node, a group of client nodes, a client file space, or a group of file spaces. You can set collocation for each sequential-access storage pool when you define or update the pool.

Figure 1 shows an example of collocation by client node with three clients, each having a separate volume that contains that client's data.

Figure 1. Example of collocation enabled by node

The server takes the data from three clients and stores each client's data on a separate tape.

Figure 2 shows an example of collocation by group of client nodes. Three groups are defined, and the data for each group is stored on separate volumes.

Figure 2. Example of collocation enabled by node collocation group

There are three groups of client nodes. One group consists of two nodes; another group consists of three nodes; the third group consists of four nodes. The server takes the data from the three groups and stores each group's data on separate tapes.

Figure 3 shows an example of collocation by file space group. Six groups are defined. Each group contains data from file spaces that belong to a single node. The data for each group is stored on a separate volume.

Figure 3. Example of collocation enabled by file space collocation group

There are six groups of file spaces. Each group contains data from file spaces that belong to a single node. The server takes the data from the six groups and stores each group's data on separate tapes.

When collocation is disabled, the server tries to use all available space on each volume before it selects a new volume. While this process provides better use of individual volumes, user files can become scattered across many volumes. Figure 4 shows an example of collocation that is disabled, with three clients that share space on single volume.

Figure 4. Example of collocation disabled

The server takes the data from three client nodes and stores all the data on one tape. The data for each client is not collocated together on the tape. It is interspersed. Two files from client one are followed by a file from client two and a file from client three. This is followed by a file from client one, two files from client two, and a file from client three.

With collocation disabled, more media mount operations might be required to mount volumes when users restore, retrieve, or recall many files.

Collocation by group is the IBM Spectrum Protect™ system default for primary sequential-access storage pools. The default for copy storage pools is no collocation.