Coming soon to FlexRAID Basic...

1. Automatic data management (storage pooling)

I will soon be working on an add-on for FlexRAID Basic that will add automatic data management.

Basically, all your data including network data will be presented as a single view to which your will write and read from.
You won't have to worry about managing your free space whether you are using the space internally or externally (through the network).

You will also be able to bypass the view without jeopardizing the RAID (FlexRAID will automatically resynch).
The one catch is, you will need an additional drive that will serve as landing drive when writing through the view.
The good news is the landing drive can be your OS drive as long as it has enough space to hold data for a short interval (something that you will configure - anywhere from 5 minutes to a few hours).

2. An "undo" feature

The next feature I will be adding is the ability to undo a failed rsynch operation.
Right now, a failed rsynch could leave the RAID in an inconsistent state.
This isn't usually an issue since you could retry the rsynch operation to fix everything.

The only situation where one could not complete an rsynch or would not want to complete it and want to undo is if a drive fails in the middle of an rsynch.
At that point, it would be good to undo the failed rsynch so you can recover the failed drive.

Being able to undo a failed rsynch requires additional parity space.
Ideally, the space required has to be equal to the amount of data being changed.
For most of us, that's only a few Gigabytes.
However, if you are adding a new drive with hundreds of Gigabytes... well, hundreds of Gigabytes of free space will be required.
That said, there is no reason to be strict with the space requirement.
So, if you are making 100GB worth of changes and only have 10GB for additional parity undo space, then I can be flexible and allow for a up to 10GB undo operation.
The undo parity DRU can contain any path you choose. So, you can spread it across multiple drives (including data, parity, and landing drives).
The undo operations should also be fairly fast with constant speed equal to a data copy operation.

Again, both mentioned features will be optional.
However, I will be switching off my focus from FlexRAID Live! to work on them.
I am running into performance issues with FlexRAID Live!, so I need to switch focus for a bit.

What do you guys think?

Brahim wrote:Basically, all your data including network data will be presented as a single view to which your will write and read from.

I'm not sure i fully understand this. Will it be like a virtual directory with access to all the drives shown with dir-names or am I completely off here?

dvdmonster wrote:

Brahim wrote:Basically, all your data including network data will be presented as a single view to which your will write and read from.

I'm not sure i fully understand this. Will it be like a virtual directory with access to all the drives shown with dir-names or am I completely off here?

Yeah, it will be like a virtual directory.
In fact, it will be like DFS with automatic data management.

Brahim wrote:
Yeah, it will be like a virtual directory.
In fact, it will be like DFS with automatic data management.

Sounds sweet.. can't wait to try it out.

FYI, I am currently working on number 2.
It's a lot more tricky than I though it would be.
I hope to be done by the end of this week.

Number 1 will follow after the release of FlexRAID 1.1.

Wow! #1 sounds like you're providing one of the main features (aka selling points) of Windows Home Server. That's pretty impressive. I believe it also gives people another option when WHS' drive pooling isn't working for them. They could leave the drives out of the pool and use this method instead.