Data path architecture

The data path of IBM Storage Scale DAS comprises three tiers that are parallel to the three tiers of the infrastructure architecture. For more information, see Infrastructure architecture.

Figure 1 illustrates how object data is represented at each layer. For information on how to configure the different layers, see Installing.

Figure 1. Example IBM Storage Scale DAS data path

The IBM Storage Scale file system for IBM Storage Scale DAS comprises one or more storage pools that contain the disks provided by the storage devices. The example illustrates one IBM Storage Scale file system (fs1 mounted at /data/fs1), which comprises two ESS based storage pools. The file system includes the directory project1-data and the file message.

ls /data/fs1/project1-data
message

cat /data/fs1/project1-data/message
IBM Storage Scale provides scalable performance.

The IBM Storage Scale DAS cluster includes IBM Storage Scale container native. IBM Storage Scale container native remotely mounts the file system fs1 at /mnt/fs1. The directory project1-data and the file message are available under the respective paths.

ls /mnt/fs1/project1-data
message

cat /mnt/fs1/project1-data/message
IBM Storage Scale provides scalable performance.

The NooBaa endpoint pods of Red Hat OpenShift Data Foundation provide S3 access to data in IBM Storage Scale. IBM Storage Scale DAS uses IBM Storage Scale CSI to make IBM Storage Scale file systems available in NooBaa endpoint pods. NooBaa mounts the IBM Storage Scale file systems in sub-directories of directory /nsfs. In this example, the file system fs1 is mounted at /nsfs/noobaa-s3res-4142975866. The directory project1-data and the file message are available under the respective paths.

ls /nsfs/noobaa-s3res-4142975866/project1-data
message

cat /nsfs/noobaa-s3res-4142975866/project1-data/message
IBM Storage Scale provides scalable performance.

IBM Storage Scale DAS makes configurable directories in IBM Storage Scale file systems accessible as S3 buckets. In this example, the directory project1-data is exported as the S3 bucket project1-bucket. The mmdas command can report all exported directories and the mapping of S3 buckets to file system directories.

mmdas export list project1-bucket

 Name                   Filesystem Path
 -----                  -----------------
 project1-bucket        /mnt/fs1/project1-data/

mmdas export list

 Name
 ------
 project1-bucket
 project2-bucket
 shared-bucket

S3 applications can access such exported directories and files as S3 buckets and S3 objects. In this example, the file message in the directory project1-data is accessible as S3 object message in the S3 bucket project1-bucket. In the following output, the command s3p1 is an alias for the AWS CLI.

Note: To set the alias for s3p1, see Example configuration of IBM Storage Scale DAS.

s3p1 ls s3://project1-bucket
2022-03-12 08:40:28         50 message

s3p1 cp s3://project1-bucket/message mymessage
download: s3://project1-bucket/message to ./mymessage

cat mymessage
IBM Storage Scale provides scalable performance.