Thanks for showing your great support of DrivePool.
I just wanted to let everyone know that DrivePool is moving forward and very cool things are in the works.
Let me give you a rough road map of what's coming:
DrivePool 1.1
I believe that the current BETA merits a minor version bump, so it will be released as 1.1.0.xxxx. This is mainly because CoveFS received a significant overhaul. Other new stuff includes fast drive removal, network I/O boost, and an updated read striping algorithm using blocks.
DrivePool 1.2
I'm looking forward to this build
The seeds for it are already getting put in place.
We will have a completely extensible file balancing / file placement plug-in system. It will be extremely simple for anyone with basic programming knowledge to write DrivePool plug-ins that control file placement and balancing.
The plug-ins will be written in .NET.
What does this mean?
For one thing it will let the Scanner do things like automatically migrate your files off damaged disks. It will also give every DrivePool user full control over file balancing. You will be able to set limits on which disks can hold duplicated vs. unduplicated files and how much.
We'll have a few built-in plug-ins, but anyone is free to write any kind of plug-in that they like.
The new plug-in based system will be layered, so you can have multiple plug-ins interacting together to build one balancing model. So you won't be limited to selecting only one balancing algorithm.
What else can be done with this?
A performance optimization plug-in. Let your fastest disks save all the new files, and then you migrate those files to the slower disks at night automatically. You can even use a SSD for this.
Prioritize disk usage by age. Perhaps you don't want to risk putting any unduplicated files on a disk that's more than 3 years old.
Heat management. Place files on the coolest disk, up to a fill limit.
These are just some ideas that sound cool to me. Anyone would be able to write these kinds of plug-ins.
DrivePool 1.3
Multiple pool support.
Yes, you've asked for it and I think we can make it happen with almost no additional UI complexity.
DrivePool 2.0
Major new pooling capabilities that are being worked out now.
----
So that's the road map, and obviously as we move into the future things may change.
Comments
Imagine you have a pool of slow but high-capacity low-speed harddisks (like WD20EARS). Those disks throttle the data transfer speed to/from the server. Then you could add some low-capacity high-speed harddisks (like WD Raptor). When transfering data to the server, the plugin would redirect the data onto the high-speed disks, so that the transfer is at maximum available speed. Later when duplication is being started, the plugin would duplicate those new files onto the slower high-capacity disks - and remove them from the high-speed disks (making room for new files). A detail could be like holding often used server-files on high-speed disks, if the user decides, so also reading from the server can reach maximum speed.
It is just around the corner so I hope something is ready prior to its official release.
Personally I think it is/was a great idea by Microsoft to create one ring to rule them all.
I agree I learned a long time ago that software raids just doesn't work and calling it storage spaces won't make it work any better. I've lost about 2tb worth of data and none of the raid reconstruction tools help at all. Hell I spent about 200 bucks trying to salvage my data.
My option is drive pool or nothing. I plan on converting my WHS2011 into a W8 server-like PC but another dilemma I am facing besides drive pool compatibility is a network full image backup solution that works similar to WHS2011 backup. Do you know of any such software?
Come on Alex the natives are restless, lol.
Di