Recalling a Volume to file system
Explains how to recall offloaded data to the file system.
About this task
When you:
- Read data offloaded to a remote target (or cold tier), data is automatically recalled to the file system to allow the read to succeed.
- Modify data offloaded to an erasure coded volume (or warm tier) or a remote target (or cold tier), data is automatically recalled to the file system to allow the write to succeed.
The recalled data is automatically:
- Purged based on the expiration time period set at the volume level for recalled data if there are no changes (for example, read operation).
- Offloaded based on the rule and the expiration time period set at the volume level for recalled data if there are changes (for example, overwrite operation).
See Data Reads, Writes, and Recalls for more information. If the recall fails, CLDB retries the operation after some time. See Retrying Failed Operation for more information.
You can manually recall all data in a volume using the Control System, CLI, or the REST API.
NOTE
For a tiered volume, there can be only one running job at any given time. If you
trigger a recall operation when another job is running on the tiered volume, the
recall operation will fail.You can also recall individual files from the tier. See Recalling a File to file system Using the CLI and REST API for more information.
Recalling Offloaded Volume Using the Control System
Procedure
Recalling Offloaded Volume Using the CLI and REST API
About this task
Run the following command to recall volume data:
/opt/mapr/bin/maprcli volume recall -name <volName>Send a request of type POST to URL. For example:
curl -k -X POST 'https://abc.sj.us:8443/rest/volume/recall?name=volName' --user mapr:maprFor more information, see volume recall.