Hi I know ZFS was designed to work as a "you do not need to know FS" and for JBOD disks, but sometimes we need to know where we have redundancies and sometimes we do use HDS or EMC arrays. I have a large costumer that needs to ensure redundancies between sites. This means ensuring that data from lun XYZ on HDS in site A is redundant on HDS in site B. 1. Is there any way to tell zfs about enclosures and policies besides just remembering to add lun A (Site A) as a mirror of lun A (Site B). It would be nice to be able to mandate such a policy to a pool and also to define the redundancies on the level of a luns on HDS A port A,B,C mirror to HDS B port A,B,C. 2. Sometimes HDS storage arrays are EOL and we need to move a large amount of luns from on array to another array. The easy fix has always been to evac the disks or create a mirror. Think of it as a SVC feature, where you just tell zfs to migrate the box. zpool migratedisks -FromSite SiteA -FromEnclousure HDSA -ToSite SiteA -ToEnclousure HDSB -policy "DoNotCopyInPeekHours" 3. When working in TrueCopy or SRDF environments it would be nice if zfs could be mandated to start a scrub when a pool is imported (the true copy or SRDF may have been corruped by the array). If these features are not on the roadmap, then I would like to know if I am the only one that needs these features from zfs? -- // Flemming Danielsen -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20070914/996db363/attachment.html>