Capacity Optimization - Start Here

CLARiiON Block Data Compression is a new capability that  allows you to save and reclaim space anywhere in your production  environment with no restrictions. Key facts about CLARiiON Block Data  Compression:

  • Data compression works as a background task to minimize performance overhead
  • Supports thin LUNs, and automatically migrates thick LUNs to thin during compression, freeing valuable storage capacity
  • Supports compression of LUNs in CLARiiON CX4 and block-enabled Celerra NS Series platforms
  • Ideal for managing proliferation of user-generated, unstructured data (text, office files, MP3/MP4, etc.)

 

Key Resources

White Papers

 

Demos


-------------------------------------------------------------------------------------------------------------------------------------------------

 

Celerra Data Deduplication intelligently reduces space usage through a combination file-level deduplication with built-in compression to deliver the maximum storage savings with the least amount of resource footprint.

 

Celerra performs all deduplication processing as a background, asynchronous operation that acts on file data after it has been written into the file system. it does not process active data or data as it is written into the file system. It avoids active data because active data is more likely to be accessed, modified, or deleted in a short time period. Inactive data, which represents the largest component of your datasets - roughly 80 to 90 percent - is targeted for the deduplication process. However, there is the ability to target selective "active" files for processing.

 

Celerra Data Deduplication also supports compression for virtual machines through the Celerra Plug-in for VMware. Not only does the deduplication process run automatically and select candidate data based on certain parameters, but end users can also manually select files for deduplication. This is achieved through Microsoft windows Explorer integration - an end user can choose a file or directory as a candidate for compression by simply selecting "Compression" on the advanced tab of the file/directory properties folder. This gives the users the control of what additional files should be stored more efficiently.

 

The system is designed to process only those files that are not being actively used by clients, thereby both maximizing the space savings and minimizing the impact on end users and applications. By default, the system selects files based on their size (minimum and maximum) and age (access and modification time). The administrator can tune these selection criteria if desired and optionally add filters to exclude specific file types (e.g. by extension/directory).

 

By running all the processing in the background and by avoiding processing active files, you avoid introducing a performance penalty on the data with which you are running your business.  Deduplication activity is also throttled to avoid impact on processes serving client I/Os. When running on an X-Blade, deduplication will process one file system at a time and throttle its activity if the X-Blade CPU utilization exceeds 75%. This means that Celerra Data Deduplication will process the bulk of the data in a file system in the background without impacting the production workload by using otherwise idle CPU cycles.


Once candidate files have been identified for deduplication, two activities take place:
  • Compression: Compression is accomplished by using technology from the EMC RecoverPoint compression engine. As discussed earlier, compression is where the bulk of the savings occur. A file that has been identified for the deduplication process will benefit from the compression process even if it is not a duplicate file.
  • Deduplication: File-level deduplication is accomplished by using technology from the EMC Avamar hashing algorithm. This is where all duplicated copies of files are identified and reduced to one instance.

 

Key Resources

White Papers

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/White_Paper/h8045-data-compression-wp.pdf

 

Demos

Demo: EMC Virtual Provisioning and Block Data Compression