It seems the introduction of EV8 has introduced some new challenges for object level recovery. I'm looking for some information on new recovery methodologies under EV8. The environments I deal with implement EV for storage optimization purposes and not compliance. Unfortunately, most allow users to delete from the archive, and then they also allow the users to request something they deleted be recovered (as far back as two years).
Prior to version 8, the audit log could be utilized to recover deleted items, either through:
- reconstructing the path to an item which remains in the dumpster in a browser window, or
- by allowing for specific DVS files to be recovered from tape (either directly accessing the DVS file through Outlook, or by mounting them onto an EV server with a db from the same timeperiod)
(these methods, while not fully sanctioned, are referenced directly in several Symantec KB articles)
While the first should continue to function, it seems the later will not due to the new OSIS engine
- Is their a way to audit, or track the assocaited deletion of object parts in v8, when the parent object is deleted, whether or not it results in the actual sub-objects deletion (to allow for selective object restores to continue to function as it did in previous releases)?
My second issue relates more to the "official" recovery process:
- prior to version 8, all data related to a deleted object existed in a single partition. Recovering deleted objects could be restricted to recovering a single partition, and in the worst case an entire store. Recovering a single partition to an alternate server for recovery purposes, while cumbersome was possible.
- version 8 introduces some new challenges, the new OSIS engine could spread data for a single object across multiple stores, multiple servers and even multiple partitions. With the amount of data most customer have, or will accumulate in their back-end vault store - it no longer becomes reasonable to expect a customer to be able to restore all vault stores, servers and partitions, to an alternate or recovery environment in order to continue performing object level recovery.
- Yes, storage groups could be broken down into reasonable size repositories, but that essentially flips us back to the pre-v8 storage architecture.
- Unfortunately, the EV dumpster is limited to 99 days (and while completely useless to most, tied to the utilization of the audit log and the methods mentioned above, it has had its benefits)
I'm very curious what is currently being recommended for recovery under the new architecture? What tools are available, or will be made avilable to assist with this process?
(and don't recommend the various environments I work in change their policies, as it is not going to happen - the business rule for object recovery was set long before EV came along, and they don't want a different rule for the storage optimized data. They currently perform these operations on their production systems and they want it the same for the EV repository.)
I'm running into this issue more and more...
Any feedback is appreciated,
Did you already had your think about this one? I can imagine this being a difficult one, but I'd like to see the answer/suggestions for this.
Thanks, take your time, don't forget it though..