I'm probably not alone when I say that I figured HDS' zero page reclaim was something more than it actually is. After I posted my bloat/shrink video a couple days ago, Hu Yoshida at HDS posted a pretty clear entry explaining that it is a fat to thin migration function. My apologies for not getting it right - I should have - and so I try in this video below, once again, to explain what it is.
Zero page reclaim has been one of the stranger announcements this year - and it certainly doesn't top Storage Anarchist's announcement coup of IBM's XIV last year, but if there were awards, for this sort of thing there might be a category for zero page reclaim this year.
Hi Marc,
Zero Page Reclaim is not a migration only feature. The example provided by HDS was to show that ZPR could be used to reclaim space back from a fat volume after it has been migrated. It can also continue to be run at the storage administrators discretion to reclaim space from volumes that have bloated reclaiming available space – even if it is in 42MB chunks.
How do you define bloat? Beyond normal growth, aren’t the biggest bloat drivers a result of thin-unfriendly file systems and inappropriate practices for thin provisioned volumes?
Cheers,
Sim
Posted by: Sim Alam | October 06, 2009 at 04:37 PM