Intelligent tiers of backup data
The claimed subject matter relates to systems and/or methodologies that facilitate intelligent distribution of backup information across storage locations in network-based backup architectures. A virtual layering of backup information across storage locations in the backup architecture can be implemented. Statistical models are utilized to dynamically re-allocate backup information among storage locations and/or layers to ensure availability of data, minimum latency upon restore, and minimum bandwidth utilization upon restore. In addition, heuristics or machine learning techniques can be applied to proactively detect failures or other changes in storage locations such that backup information can be reallocated accordingly prior to a failure.
1. A system that facilitates intelligent allocation of backup data among a set of storage locations in a hybrid backup environment, comprising:
a processor coupled to a memory that retains computer-executable instructions, the processor executes:
a monitor component that identifies properties of backup data stored by one or more cloud storage locations and one or more peer-to-peer storage locations, and properties of the one or more cloud storage locations and the one or more peer-to-peer storage locations; and
a tier component that implements virtual layers of backup data across the one or more cloud storage locations and the one or more peer-to-peer storage locations, the one or more cloud storage locations being remote from the one or more peer-to-peer storage locations, in accordance with the properties of the backup data and the properties of the one or more cloud storage locations and the one or more peer-to-peer storage locations, the tier component distributes backup data among the one or more cloud storage locations and the one or more peer-to-peer storage locations to ensure availability while reducing storage utilization and latency upon restore of the backup information, wherein frequency of access to the backup data of the one or more cloud storage locations and the one or more peer-to-peer storage locations is utilized to distribute the backup data among the one or more cloud storage locations and the one or more peer-to-peer storage locations, and
wherein the tier component distributes backup data among each of the one or more cloud storage locations and the one or more peer-to-peer storage locations by dividing a file into a plurality of segments and distributing a first portion of the plurality of segments to the one or more cloud storage locations and a second portion of the plurality of segments to the one or more peer-to-peer storage locations, remote from the one or more cloud storage locations.
2. The system of claim 1 , wherein the monitor component includes a data evaluation component that analyzes backup data to ascertain the properties of the backup data.
3. The system of claim 1 , wherein the monitor component includes a machine evaluation component that observes at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations to determine the properties of the one or more cloud storage locations and the one or more peer-to-peer storage locations.
4. The system of claim 1 , wherein the tier component includes a distribution component that replicates a block of backup data to at least one of the one or more cloud storage locations and/or the one or more peer-to-peer storage locations based at least in part on the properties of the backup data or the one or more cloud storage locations and the one or more peer-to-peer storage locations.
5. The system of claim 1 , wherein the tier component includes an indexing component that maintains an index, the indexing component at least one of adds, deletes, or modifies entries in the index when distribution decisions are rendered by the tier component.
6. The system of claim 5 , wherein the index comprises a listing of relationships between backup versions and at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations to which the backup versions have been distributed.
7. The system of claim 1 , wherein the properties of the backup data include at least one of frequency of access to the backup data, availability of the backup data, or time since creation of the backup data.
8. The system of claim 7 , wherein frequently accessed backup data is inferred to be most likely to be restored, and wherein backup data is frequently accessed when the backup data is accessed a predetermined number of times within a specific time period.
9. The system of claim 8 , wherein the tier component distributes frequently accessed data to at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations having minimal latency and highest availability.
10. The system of claim 8 , wherein the tier component replicates copies of frequently accessed backup data to at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations.
11. The system of claim 7 , wherein infrequently accessed backup data is inferred to be least likely to be restored, and wherein backup data is infrequently accessed when the backup data is accessed less than a predetermined number of times within a specific time period.
12. The system of claim 11 , wherein the tier component allocates infrequently accessed backup data to remote storage nodes.
13. The system of claim 1 , wherein the properties of at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations comprises health of each respective one or more cloud storage locations and one or more peer-to-peer storage locations, storage capacity of each respective one or more cloud storage locations and/or the one or more peer-to-peer storage locations, availability of each respective one or more cloud storage locations and/or the one or more peer-to-peer storage locations, bandwidth utilization of each respective one or more cloud storage locations and/or the one or more peer-to-peer storage locations, or predicted latency times for transmission of data between each respective the one or more cloud storage locations and/or the one or more peer-to-peer storage locations.
14. The system of claim 1 , wherein the tier component detects a probability of a failure of at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations or a client machine based upon the properties of at least one of the one or more cloud storage locations and the one or more peer-to-peer storage locations.
15. The system of claim 14 , wherein the tier component proactively allocates backup data prior to an occurrence of the failure.
16. The system of claim 1 , wherein the tier component creates the virtual layers of the backup data through application of a higher level of preference on storage locations corresponding to the one or more peer-to-peer storage locations than to storage locations corresponding to the one or more cloud storage locations.
17. A method for intelligently tiering backup information in a distributed hybrid backup environment, comprising: employing a processor executing computer-executable instructions stored on computer-readable storage medium to implement the following acts:
creating virtual layers of backup information across one or more peer-to-peer storage locations and one or more cloud storage locations of the hybrid backup environment, the one or more peer-to-peer storage locations being remote from the one or more cloud storage locations;
monitoring backup information to ascertain properties of the backup information, the properties including each of access frequency, availability, and time since creation of the backup information; and
dynamically reallocating backup information across each of the one or more cloud storage locations and the one or more peer-to-peer storage locations, based upon the properties of the backup information, to ensure availability of the backup information while minimizing storage costs and latency upon restoration of the backup information,
wherein the reallocation includes moving backup information accessed less than a predetermined number of times within a specific time period from the one or more peer-to-peer storage locations to the one or more cloud storage locations during off-peak times and wherein the backup information is reallocated among each of the one or more cloud storage locations and the one or more peer-to-peer storage locations by dividing a file into a plurality of segments and distributing a first portion of the plurality of segments to the one or more cloud storage locations and a second portion of the plurality of segments to the one or more peer-to-peer storage locations remote from the cloud storage location.
18. The method of claim 17 , further comprising:
designating the backup information as at least one of hot data or cold data based upon the properties of the backup information, wherein hot data is backup information that is accessed a predetermined number of times within a specific time period and cold data is backup information that is accessed less than the predetermined number of times within the specific time period;
allocating hot data to storage locations that provide optimal locality to a restore client, wherein optimal locality is provided by storing the hot data in a storage location that is close on the network to the restore client; and
allocating cold data to remote storage locations that provide storage at minimum cost.