Autonomous data protection
Embodiments of the present disclosure relate to autonomous data protection. One or more input/output (I/O) streams can be received by one or more storage devices. One or more snapshots of each storage device can be obtained. One or more anomalies can be identified based on a change rate corresponding to each storage device's memory allocated to store user data included in the I/O streams.
1. An apparatus comprising a memory and at least one processor configured to:
receive one or more input/output (I/O) streams by one or more storage devices;
obtain one or more snapshots of each storage device;
identify one or more anomalies based on a change rate corresponding to each storage device's memory allocated to store user data included in the I/O streams;
allocate a portion of each storage device's memory to store each snapshot, wherein the portion includes one or more of: an anomaly sub-portion, a continuous snaps sub-portion, and a free space sub-portion;
store each snapshot in the continuous snaps sub-portion; and
store each identified anomaly in the anomaly sub-portion.
2. The apparatus of claim 1 further configured to obtain the one or more snapshots of each component of each storage device.
3. The apparatus of claim 1 further configured to obtain the one or more snapshots based on one or more of a predetermined snapshot rate and a dynamically determined snapshot rate.
4. The apparatus of claim 3 further configured to present the allocated portion to each user of each storage device as free space.
5. The apparatus of claim 1 further configured to:
monitor each storage device's available user data memory configured to store the user data; and
dynamically allocate additional memory to the available user data memory from the free space sub-portion based on one or more of a current amount of available user data memory and predicted amount of available user data memory.
6. The apparatus of claim 1 further configured to:
purge data stored from one or more of the sub-portions based on one or more predetermined events; and
prevent data flagged as protected by a user from being purged.
7. The apparatus of claim 1 further configured to identify each anomaly based on a change rate of changed tracks of each logical unit number (LUN) of each storage device.
8. The apparatus of claim 1 further configured to:
for each anomaly remaining in an unconfirmed state after a predetermined period, flag the anomaly as a false positive; and
generate an anomaly detection model based on confirmed anomaly data, wherein generating the anomaly detection model includes providing a machine learning (ML) engine that processes the confirmed anomaly data.
9. The apparatus of claim 8 , wherein the ML engine is configured to apply a reinforcement learning technique on the confirmed anomaly data to generate the anomaly detection model.
10. A method comprising:
receiving one or more input/output (I/O) streams by one or more storage devices;
obtaining one or more snapshots of each storage device;
identify one or more anomalies based on a change rate corresponding to each storage device's memory allocated to store user data included in the I/O streams;
allocating a portion of each storage device's memory to store each snapshot, wherein the portion includes one or more of: an anomaly sub-portion, a continuous snaps sub-portion, and a free space sub-portion;
storing each snapshot in the continuous snaps sub-portion; and
storing each identified anomaly in the anomaly sub-portion.
11. The method of claim 10 further comprising obtaining the one or more snapshots of each component of each storage device.
12. The method of claim 10 further comprising obtaining the one or more snapshots based on one or more of a predetermined snapshot rate and a dynamically determined snapshot rate.
13. The method of claim 12 further comprising presenting the allocated portion to each user of each storage device as free space.
14. The method of claim 10 further comprising:
monitoring each storage device's available user data memory configured to store the user data; and
dynamically allocating additional memory to the available user data memory from the free space sub-portion based on one or more of a current amount of available user data memory and predicted amount of available user data memory.
15. The method of claim 10 further comprising:
purging data stored from one or more of the sub-portions based on one or more predetermined events; and
preventing data flagged as protected by a user from being purged.
16. The method of claim 10 further comprising identifying each anomaly based on a change rate of changed tracks of each logical unit number (LUN) of each storage device.
17. The method of claim 10 further comprising:
for each anomaly remaining in an unconfirmed state after a predetermined period, flagging the anomaly as a false positive; and
generating an anomaly detection model based on confirmed anomaly data, wherein generating the anomaly detection model includes providing a machine learning (ML) engine that processes the confirmed anomaly data.
18. The method of claim 17 further comprising applying a reinforcement learning technique on the confirmed anomaly data using the ML engine to generate the anomaly detection model.