Check out the below links for info direct from HDS on the Zero Page Reclaim feature that works with Dynamic Provisioning as of ucode 60-04-04 on the USP V -
Webex playback – requires registration
PDF of the PPT accompanying the above Webex session -
If the above pages move, you can see all publically available WebTech sessions from www.hds.com/webtech
On the Webex session there is also mention of the anticipated availability of being able to dynamically re-balance page distribution within a Pool. This is another feature that lots of people have talked to me about, basically when adding more capacity to an existing Pool will the Page layout be re-balanced (levelled) over all of the existing AND new capacity. The answer at the moment is No, but it sounds increasingly like this feature will hit the streets later this year.
Hopefully this may lead to the ability to migrate at the Page level via HTSM? Pure speculation on my behalf.
Quick question… can you run ZERO Out feature on live production environment? Will this added any additional overhead ?
Jay,
The Zero Page Reclaim on the USP V/XP24000 (I think its actually called Discard Zero Data in the GUI) is run as a background task with a low priority so as to have minimum impact on host IO. Kind of similar to a background Shadow Image copy job or a sparing out operation, only at a far lower priority. You should not notice a performance hit during the operation.
Just to be sure, I assume that when you say "ZERO Out" you are referring to the Zero Page Reclaim operation performed on the array, and not the way in which you use a host based tool to write zeros to deleted areas of a filesystem (a nifty way to increase the amount of free space you can claim back with a Zero Page Reclaim is to first Zero Out, usually with a secure erase tool, all previously used but now deleted space within a filesystem). I will probably post on this in the future if you are unsure or there is interest.
Nigel
PS. Glad you like the blog