Measurement interval anomaly detection-based generation of snapshots
An illustrative method includes a data protection system determining a metric associated with operations performed with respect to a storage system during a measurement interval, determining that the metric deviates by more than a threshold amount from a historical baseline metric associated with the storage system, and directing, based on the determining that the metric deviates by more than the threshold amount from the historical baseline metric, the storage system to generate a recovery dataset for data maintained by the storage system.
1. A method comprising:
determining, by a data protection system, a metric associated with operations performed with respect to a storage system during a measurement interval;
determining, by the data protection system, that the metric deviates by more than a threshold amount from a historical baseline metric associated with the storage system;
directing, by the data protection system based on the determining that the metric deviates by more than the threshold amount from the historical baseline metric, the storage system to generate a recovery dataset for data maintained by the storage system, the recovery dataset initially classified as a provisional recovery dataset that has a first set of criteria that has to be met for the recovery dataset to be deleted;
determining, by the data protection system, that an additional metric associated with operations performed with respect to the storage system during one or more measurement intervals subsequent to the measurement interval also deviates from the historical baseline metric by more than the threshold amount; and
reclassifying, by the data protection system based on the determining that the additional metric also deviates from the historical baseline metric by more than the threshold amount, the recovery dataset as a protected recovery dataset, the protected recovery dataset having a second set of criteria that has to be met for the recovery dataset to be deleted, the second set of criteria more stringent than the first set of criteria.
2. The method of claim 1 , further comprising logging, by the data protection system, data instances affected by the operations performed during the measurement interval.
3. The method of claim 2 , further comprising performing, by the data protection system, a data recovery process with respect to one or more of the logged data instances.
4. The method of claim 3 , further comprising:
logging, by the data protection system based on the determining that the metric deviates by more than the threshold amount, additional data instances affected by additional operations performed during a time period subsequent to the measurement interval;
wherein the performing the data recovery process is further with respect to one or more of the additional data instances.
5. The method of claim 2 , wherein the data instances comprise at least one of one or more logical block addresses, one or more files, or one or more objects.
6. The method of claim 2 , further comprising determining, by the data protection system based on the logging, whether the data maintained by the storage system is possibly being targeted by a security threat.
7. The method of claim 1 , wherein:
the operations comprise write operations that write data to the storage system during the measurement interval;
the metric comprises an entropy metric associated with the measurement interval, the entropy metric indicating a percentage of data written to the storage system during the measurement interval that has an entropy greater than an entropy threshold;
the historical baseline metric comprises a historical entropy metric associated a historical baseline interval prior to the measurement interval, the historical entropy metric indicating a percentage of data written to the storage system during the historical baseline interval that has an entropy greater than the entropy threshold.
8. The method of claim 7 , wherein the determining that the metric deviates by more than the threshold amount from the historical baseline metric comprises determining that the entropy metric is greater than the historical entropy metric by more than the threshold amount.
9. The method of claim 1 , further comprising dynamically determining, by the data protection system, the threshold amount based on an amount of data associated with the historical baseline metric.
10. The method of claim 1 , wherein the determining that the metric deviates from the historical baseline metric by more than the threshold amount comprises determining that the metric deviates from the historical baseline metric by more than a predetermined number of standard deviations.
11. The method of claim 1 , further comprising using, by the data protection system, the recovery dataset to perform a data recovery process with respect to the data maintained by the storage system.
12. The method of claim 1 , wherein the measurement interval has a predetermined time duration.
13. The method of claim 1 , wherein the measurement interval is a rolling time interval.
14. The method of claim 1 , wherein the measurement interval corresponds to a time period during which a predetermined number of operations is performed with respect to the storage system.
15. The method of claim 1 , wherein a duration of the measurement interval is based on an amount of log data acquired by the data protection system.
16. The method of claim 1 , wherein the recovery dataset corresponds to a point in time that is prior to the measurement interval.
17. A system comprising:
a memory storing instructions; and
a processor communicatively coupled to the memory and configured to execute the instructions to:
determine a metric associated with operations performed with respect to a storage system during a measurement interval;
determine that the metric deviates by more than a threshold amount from a historical baseline metric associated with the storage system;
direct, based on the determining that the metric deviates by more than the threshold amount from the historical baseline metric, the storage system to generate a recovery dataset for data maintained by the storage system, the recovery dataset initially classified as a provisional recovery dataset that has a first set of criteria that has to be met for the recovery dataset to be deleted;
determine that an additional metric associated with operations performed with respect to the storage system during one or more measurement intervals subsequent to the measurement interval also deviates from the historical baseline metric by more than the threshold amount; and
reclassify, based on the determining that the additional metric also deviates from the historical baseline metric by more than the threshold amount, the recovery dataset as a protected recovery dataset, the protected recovery dataset having a second set of criteria that has to be met for the recovery dataset to be deleted, the second set of criteria more stringent than the first set of criteria.
18. A non-transitory computer-readable medium storing instructions that, when executed, direct a processor of a computing device to:
determine a metric associated with operations performed with respect to a storage system during a measurement interval;
determine that the metric deviates by more than a threshold amount from a historical baseline metric associated with the storage system;
direct, based on the determining that the metric deviates by more than the threshold amount from the historical baseline metric, the storage system to generate a recovery dataset for data maintained by the storage system, the recovery dataset initially classified as a provisional recovery dataset that has a first set of criteria that has to be met for the recovery dataset to be deleted;
determine that an additional metric associated with operations performed with respect to the storage system during one or more measurement intervals subsequent to the measurement interval also deviates from the historical baseline metric by more than the threshold amount; and
reclassify, based on the determining that the additional metric also deviates from the historical baseline metric by more than the threshold amount, the recovery dataset as a protected recovery dataset, the protected recovery dataset having a second set of criteria that has to be met for the recovery dataset to be deleted, the second set of criteria more stringent than the first set of criteria.
19. The system of claim 17 , wherein the processor is further configured to execute the instructions to log data instances affected by the operations performed during the measurement interval.
20. The system of claim 19 , wherein the processor is further configured to execute the instructions to perform a data recovery process with respect to one or more of the logged data instances.