Data placement policy
How the IBM Storage Archive uses the data placement policy.
IBM Storage Archive uses two partitions on the medium: the index partition and data partition. When you mount the medium, an index file is written to both partitions. The index file on the index partition is overwritten by the new index file. However, the index file on the data partition is appended to the end of the data. Files on the index partition, including the index file, are readable the next time the medium is mounted.
The rules option for the formatting operation sets the following policies:
- File size: smaller than xxx bytes
- File name:
*.jpg
,*.txt
, and so on
Sequence of data placement policy
The data placement policy writes in the following sequence:- When a file to write matches the rules as shown in the following example:If the policy is set to (1) a file size less than or equal to 1 MB and (2) a file name of
*.txt
and*.jpg
, then the following files match the rules:- File
abc.txt
with size 512 kB - File
xyz.jpg
with size 1 MB
- File
abc.csv
with size 512 kB - File
xyz.jpg
with size 1.2 MB - File
stu.png
with size 2 MB
- File
- IBM Storage Archive writes the file to the data partition and stores it in the memory buffer.
- When the memory buffer is full or the medium is requested to unmount, IBM Storage Archive does the following actions:
- Writes an index file at the end of the data partition
- Writes the files in the memory buffer to the index partition
- Builds an index file and writes it at the end of the index partition
At the next mount, the position of the tape head is near the files in the index partition so that they are quickly readable if those files accessed just after that. Once the files in the data partition are accessed, the access performance is equal to that of the other files in the data partition. If some specific file suffixes are frequently accessed after they are mounted, it is recommended that the user set the data placement policy to include them to improve file access performance.