Data Deduplication
Combining Source Deduplication and Target Deduplication for Maximum Storage Efficiency
AIMstor is "content aware" of the deduplication that occurs. Legacy backup deduplication solutions typically do not understand the content, and this often creates issues for restore and long term data management.
The AIMstor Unified Repository employs several schemes for optimal de-duplication both within the store, and from the originating primary data source, and makes no distinction between file or email data.
- Duplicate Transfer Avoidance
- During initial synchronization AIMstor intelligently analyzes data and will not transmit any duplicates of data stored.
- Byte Level Change Detection
- Data Change Detection eliminates continuous scans, saving time, performance and capacity hits on all machines. Only changed bytes need to be transferred.
- Multi-Level Single Instance Storage
- AIMstor Unified Repository allows only a single instance of a file, no matter if it is a snapshot, CDP, archive, file or email.
- Post Process Deduplication
- Final step, runs post processing deduplication algorithms within a repository across all data sets from all machines.
- Data Reduction via Classification
- By Classifying data in real-time, AIMstor can optionally target highly specific data for any policy.
AIMstor 4-tier backup deduplication provides the optimum data reduction techniques surpassing every other backup and recovery product. Maximum performance with minimum latency is achieved at every level.
Source Deduplication is a method of reducing data that is copied and transferred at the primary data source, compromised of two technologies.
Duplicate Transfer Avoidance: When AIMstor is fist installed and setup for any type of data movement, an initial synchronization occurs so that the AIMstor Repository (data destination) and Node (source machine) are matched. The benefit is that afterwards, this synchronization no longer needs to be done again, and the AIMstor Node and Repository now have an intelligent understanding of each other. Files that are already within the Repository, do not need to be transferred from the Node, saving valuable bandwidth.
Byte-Level Change Detection: Also referred to as "Changed Byte Transfer", the AIMstor node (by virtue of AIMstor's internal Data Change Engine) is aware of what bytes have changed, without the need for continuous scans, and sends only the changes made to files and application data, such as databases, flat files and indices.
Between multiple snapshots, only the changed bytes that have taken place are sent to the Repository. Unlike a traditional backup solution, where the complete data set is copied and then needs to be de-duplicated, AIMstor inherently de-duplicates data on the fly by not unnecessarily duplicating data in the first place. Preventing data duplication at the forefront is substantially more efficient in both processing power and data stored compared to a post-operation deduplication process.
Data Reduction via Classification: Because AIMstor has the ability to classify data in real-time, any policy can include broad or granular data classifications, enabling the user to filter out data on the fly. If you have included or excluded certain files based on affiliation to group, owner, user, path, age storage type or file type, you can create highly specific policies with targeted Recovery Point Objectives and Retention for such data (i.e., only the Finance Group's Excel and PDF files, but exclude the CFO and Controller. Or, only data that is created before or after X date). See Data Classification tab for more information.
The combination of Changed Byte Transfer and Data Classification provide exponential Source Side Data Reduction capabilities that are simply not found in any solution but AIMstor.
Target Deduplication is a method of data reduction that takes place within the AIMstor Repository, reducing data from multiple sources on a global and cross-solution basis, which is separated into two distinct technologies.
Multi-Level Single Instance Store: The AIMstor Repository is a Unified Store that houses various types of protected (Backup, CDP), replicated (Synchronous, Asynchronous) and archived (File, Email) data. Because all these various forms of Repository data are stored together, AIMstor is able to "single-instance" the entire Repository, and keep track of everything in real-time through the AIMstor Metadata Store. The result is a tremendous reduction in total repository data stored.
Post Process Deduplication: AIMstor is a next generation application built upon the search engine technology, which is a big departure compared to competitive data protection and archival offerings. The result is that AIMstor can provide both the Multi-Level Single Instance Store, and Post Process Deduplication. AIMstor's Repository naturally indexes data within the repository, and eliminates duplicates that might exist in any form through a check sum post process operation.
Click box "Legacy Backup" to see inefficiencies of old backup methods.
Click box "AIMstor + Source Dedupe" to see how Cofio can eliminate over 90% increased storage capacity and network bandwidth savings.
Click box "AIMstor + Target Dedupe" to see how Cofio's Repository adds further increased efficiencies to further reduce capacity at the storage target.