Isilon - CloudPools Pros and Cons

Pros:

Can tier off to potentially cheaper storage.

Inline access for nfs/smb.

Integrated into filepool policies with the numerous condition filters.

Cons:

A secondary point of failure for files tiered off.

No option to have stubs continue to count against smartquota. (Right?)

No recall on access via smartpool jobs.

Breaks any type of metadata analysis on filesize you might have. Unsure if this is true for insightiq or not.

Danielsj1,

Thank you for sharing!

@Community,  please keep this going!

One thing that caught my attention with CloudPools is the loss of all ADS data attached to the file.  It's silently discarded.

If we were to go that route it would be with segregated data since I can't see it being intermixed with non-CloudPool data.

"No option to have stubs continue to count against smartquota. (Right?)"

This one is a real bummer for us. Imagine a policy like "move data that has not been accessed since X months on Cloudpools". The user could see its storage usage decrease over time without doing anything. The extreme case would be someone having near no apparent storage usage (besides the smartlink files) with terabytes of data on the cloudpool target.

The documention says :

"""

SmartQuota monitors and enforces administrator-defined storage limits within an Isilon cluster; it

does not enforce a limit on the amount of user or directory space used by data archived to ECS. An

archived file is only charged for the space used by the SmartLink file, thus CloudPools has the

effect of allowing user and/or directories to logically, but not physically, exceed the quota space

in SmartQuota on the Isilon cluster.

It is therefore important to be aware that for recalled files, the full file size counts against

quotas. As an example, if a 1G file is archived, only the remaining SmartLink file size is counted

against the quota. If the file is recalled, the full 1G will be counted against the quota.

As such, it is recommended to only recall files as needed and then if appropriate re-archive them so

as to not exceed quotas. Space usage can also temporarily be affected if you are caching cloud data

in SmartLink files when a user or application accesses a SmartLink file on the cluster.

"""

From the user point of view, I think it makes no sense and will generate a lot of call to the helpdesk.

Is there something on the roadmap that will offer the option to include the staged files in the quota ? It would be *very* nice to have an official feedback on this matter ?

it would be nice to have a tool to have the distribution of pool usage for a given directory (1TB is on this tier, 500G on this one and 10TB on this cloudpool target). 

Hi chjatwork,

For those interested, in the OneFS 8.1 CloudPools Administration Guide there is a section on limitations and expected behaviors that seems relevant to this thread.

https://support.emc.com/docu84274

As of the time of this post, page 52 has the information.