FlexRAID on WHS

This message was edited 1 time. Last update was at 11/10/2010 16:02:04

I wanted to post a mini-HowTo for setting up FlexRAID on WHS; or how "I" set it up at least

First, my system:

G33 Motherboard, with 6 onboard sata ports, and seperate PCIe 4port Promise controller.

Motherboard ports:
Port0 - 320GB OS/Data partitions
Port1 - 1TB Data
Port2 - 1TB Data
Port3 - 1TB Data
Port4 - 1TB Data
Port5 - 1TB Data

Promise ports:
Port6 - 1TB Data
Port7 - empty
Port8 - 750GB Data
Port9 - 1TB Parity Drive

I use my WHS system for family file storage of pictures, videos, movies, data files and PC backups. These all use the WHS duplication feature for protection.
I also have a WHS managed share for DVD movies, ripped to VIDEO_TS folders. These are protected by FlexRAID as detailed below, but still presented by WHS.

I wanted to maintain WHS's structure for presenting data shares, and for flexible, on-the-fly, data store expansion, so I didn't want to pull the DVD data out to seperate drives. But, we need to have drive level access to this data, since a drive is the unit of risk we're trying to protect.

So, on the C: root, I created a FlexRAID folder. Within this folder, I've got DATA and PARITY directories. Within each of those, I have Disk# directories, as I've assigned them: Data has Disk0, Disk1, Disk2, Disk3, Disk4, Disk5, Disk6, Disk7 (empty) and Disk8; Parity has only Disk9 at this time. These Disk# directories are hard links (created using NTFSLink http://elsdoerfer.name/=ntfslink) to the relevant DVD storage directories on each individual volume in the system, as follows:

Disk0 - D:\DE\shares\DVDs\ (this is the first Data partition in the system, and acts as the "landing zone" for WHS)
Disk1 - C:\fs\U\DE\shares\DVDs\ (the "U" represents the mount point designation for my Port1 disk, this will be different on your systems)
Disk2 - C:\fs\G\DE\shares\DVDs\ (Port2 mount point)
...
Disk9 - Partition directly mounted to here, through Disk Management (no drive letter assigned)

The above mount structure was put into place to bypass the FlexRAID restriction against hidden directories (the DE directory on each drive is hidden). I don't know if it's still required. This structure also makes it easy to create our command line scripts or WebUI configs for DRUs and Parity units.

Once the drive/directory structure is created, it's time to install software...

FlexRAID server component was installed to it's default directory.
FlexRAID client component was also installed to it's default directory.
FlexRAID WebUI was unzipped (no installer at this point), and run via the start.bat script. (No need to change any port numbers, the default 8080 is working fine for me)

That's all that's needed to get this functioning. Using the mountpoint directory structure lets us keep all of WHS in place and functioning as usual, while FlexRAID kind of works "around" it, without interferring. The WHS system is used like normal, I can copy, delete, move files to any and all shares. FlexRAID runs a rsynch every night, and corrects any of these changes I might make.

Finally, WHS "might" decide to move files around on the drives a bit, balancing things as it sees fit. All that's needed to allow for this is a recurring "rsynch" task created and run on a scheduled basis (I have this running once a night). If you decide to back up more than one share, there are a couple of ways to do this:

If you want to protect ALL the shares, merely change the directory mountpoints to point at the DE\shares\ directories, this will grab everything beneath there.
If you only want to protect a subset of shares, mount each share directory (shares\DVDs\ above) as it's own "diskunit" point. So you might have A\DE\shares\DVDs\; A\DE\shares\Music\; B\DE\shares\DVDs\; B\DE\shares\Music\; etc.

Great write up.

This is a great example of how you can/should use various techniques to protect various data types.
I too have data that I simply mirror and data that I have under FlexRAID protection.

I am curious as to why I never head of NTFSLink before.
That's a great tool for windows users that don't have DFS.

I would have thought that there was a way to tell WHS not to move certain data around.
Of course, FlexRAID has no problem dealing with data that's being moved around.
However, for efficiency sake, it would be nice to have a pool that WHS does not move around and a pool that it does what it pleases with.

Prelector:

Thanks for the information. I think this thread is great information as running WHS and FlexRaid together takes some special consideration in setting it up and potential risks running it.

I have a few questions for you.

Q 1. Is the reason you are the NTFSLink, is it to bypass the hidden directories only?

I was curious about the hidden “feature” of FlexRaid as well. I setup the DRU to point to the DE hidden directories and it backed up everything just fine. I was curious about it and found documented that only hidden FILES was not included in the parity. Since FlexRaid did create parity on the hidden \DE folders I assumed that to be correct. If that’s the case, I would assume that you don’t really need the NTFSLink program and can map DRUs to your xxx\shares\DVDs\ directories.

I was planning to use FlexRaid to backup pretty much all my shared folders. Like this:

- FlexRaid:
\Software programs
\Videos
\Books
\”PC backups”

-- WHS duplication (inside FlexRaid parity)
\Important Documents
\Pictures

After reading your post, I am now debating if I should use duplication for everything except the video folder as you are. However, I do have a share with over 40GB of stuff in it and feel it’s a bit of waste to run duplication on it. I’ll need to ponder this some more ?

Q 2. You stated in a post that you turned off Shadow Copy on all drives. Is that because it’s not really needed since you are running FlexRaid?

I looked at my drives and only D: partition had the Shadow Copy enabled. From what I can tell, it’s a feature in WHS that keeps track of copies, like a version control feature where you can get to older files.

Q 3. Can the WHS “file rebalancing” feature cause parity to be invalid? How probable is this scenario?

I was thinking about a scenario where WHS would move some files around and this would happen.

1. WHS would remove file1 from disk1 from location 1-4.
2. It would put it at disk2 location 7-10.
3. It would move another file, file2 from disk3 from location 70-71.
4. File2 would be moved to disk1 location 1-2.

In this scenario, disk1 will now have invalid entries for file1 when it comes to the parity set. If WHS were to loose any HD, except disk1, how would FlexRaid handle that situation? I would assume that file 1 would be lost, right?

olgemap wrote:
Q 1. Is the reason you are the NTFSLink, is it to bypass the hidden directories only?

I was curious about the hidden “feature” of FlexRaid as well. I setup the DRU to point to the DE hidden directories and it backed up everything just fine. I was curious about it and found documented that only hidden FILES was not included in the parity. Since FlexRaid did create parity on the hidden \DE folders I assumed that to be correct. If that’s the case, I would assume that you don’t really need the NTFSLink program and can map DRUs to your xxx\shares\DVDs\ directories.

Yep, I was only using NTFSLink as a clean pointer structure to reach the DVD directories due to the "hidden bypass" feature of FlexRAID. To be honest, I never really did any testing on that feature, and just began this setup using the old command line client. I'll probably re-examine this function when I get some time, now that the new GUI client is out.

olgemap wrote:
I was planning to use FlexRaid to backup pretty much all my shared folders. Like this:

- FlexRaid:
\Software programs
\Videos
\Books
\”PC backups”

-- WHS duplication (inside FlexRaid parity)
\Important Documents
\Pictures

After reading your post, I am now debating if I should use duplication for everything except the video folder as you are. However, I do have a share with over 40GB of stuff in it and feel it’s a bit of waste to run duplication on it. I’ll need to ponder this some more ?

With drive space as cheap as it is, I prefer to use WHS for simple duplication protection. Nice easy interface, and no need to "restore" anything, since WHS handles all that in the background. Also, remember, that the more stuff changes, the less "efficient" that FlexRAID is. FlexRAID was designed to protect ARCHIVED data, that is data that doesn't change (or not likely to change often). Music, and pictures are things that are likely to change due to metatag info, same with videos. Documents are probably being edited, etc... You either create this data as it's own protection scheme (then it's not much more efficient than WHS is), or you include it into the DVD protection scheme (in which case you're burning alot of cycles/time on rsynching). Much easier (for me at least) to just leave this to WHS to handle. Either way, I'd HIGHLY recommend you leave the "pc backups" from WHS out of any protection scheme you implement other than WHS ones... this is all database info, and as such, is very touchy about being played with. While you could probably run the parity creation without much problem, I'd be very wary of restoring them if there was any kind of problem, and not much point in backing up if you can't restore

olgemap wrote:
Q 2. You stated in a post that you turned off Shadow Copy on all drives. Is that because it’s not really needed since you are running FlexRaid?

I looked at my drives and only D: partition had the Shadow Copy enabled. From what I can tell, it’s a feature in WHS that keeps track of copies, like a version control feature where you can get to older files.

Shadowcopies are for keeping previous versions of files safe from edits/changes. But, with the "corruption bug" of WHS, you shouldn't be doing any changes in place on WHS, so this feature is wasted space.

olgemap wrote:
Q 3. Can the WHS “file rebalancing” feature cause parity to be invalid? How probable is this scenario?

I was thinking about a scenario where WHS would move some files around and this would happen.

1. WHS would remove file1 from disk1 from location 1-4.
2. It would put it at disk2 location 7-10.
3. It would move another file, file2 from disk3 from location 70-71.
4. File2 would be moved to disk1 location 1-2.

In this scenario, disk1 will now have invalid entries for file1 when it comes to the parity set. If WHS were to loose any HD, except disk1, how would FlexRaid handle that situation? I would assume that file 1 would be lost, right?

This is why I run the rsynch process every night. This process compares the parity data to the existing datasets, and updates to reflect any changes. With PP1 on WHS, there are alot fewer of these moves and changes anyways, but that rsynch process will cover any that occur (it also covers any additions or deletions I make to the datasets).

In your scenario above, you'd have to perform an rsynch of the datasets/disks after all the moves are finished. In case of a file loss/disk failure, you'd need to perform a restore on that lost disk first, then test your data integrity. There's some processes for this (verify and validate). I'd, personally, restore the lost disk, then do a verify on the entire parity set. If anything's "out of whack", then expand the scope of your restores to include the additional disks that show faulty from the verify.

Do the initial create, then run scheduled rsynch processes, and that's as close to "seamless" as we can get at this point (and it's pretty damn close).

Prelector wrote:
...
Music, and pictures are things that are likely to change due to metatag info, same with videos. Documents are probably being edited, etc...

Actually, FlexRAID will handle those just fine as each of those files won't be edited every single minute.
However, you do make a good point in mirroring what can be mirrored and RAID'ing what makes sense to be RAID'ed.

I RAID my DVDs, movies, mp3, photos, Norton Ghost backup images and system state backups, ISOs images, etc.

I mirror all of my important documents (taxes and the likes), software development workspaces (in addition to version control), etc.

So, if it is small/valuable enough to be mirrored, mirror it.
Otherwise, protect it through parity.

I personally think having WHS move the data around is a bad idea.
I just don't get why you wouldn't want to manually distribute the data yourself.

Make sure that the parity data is never moved by WHS.
You might lose protection if that happens.

Brahim wrote:

Prelector wrote:
...
Music, and pictures are things that are likely to change due to metatag info, same with videos. Documents are probably being edited, etc...

Actually, FlexRAID will handle those just fine as each of those files won't be edited every single minute.
However, you do make a good point in mirroring what can be mirrored and RAID'ing what makes sense to be RAID'ed.

I RAID my DVDs, movies, mp3, photos, Norton Ghost backup images and system state backups, ISOs images, etc.

I mirror all of my important documents (taxes and the likes), software development workspaces (in addition to version control), etc.

So, if it is small/valuable enough to be mirrored, mirror it.
Otherwise, protect it through parity.

I personally think having WHS move the data around is a bad idea.
I just don't get why you wouldn't want to manually distribute the data yourself.

Make sure that the parity data is never moved by WHS.
You might lose protection if that happens.

I don't like WHS moving data around either... but MS didn't consult me on that design decision

Seriously though, it's a backend process that we have NO control over at all. With PP1 (power pack 1), there's ALOT less of that going on, but the "potential" is still there.

And per my design detailed above, I leave the parity on it's own driver, completely seperate from WHS as a whole. It's mounted to a seperate directory on the system drive, so completely outside the WHS operations.

Prelector wrote:

... Either way, I'd HIGHLY recommend you leave the "pc backups" from WHS out of any protection scheme you implement other than WHS ones... this is all database info, and as such, is very touchy about being played with. While you could probably run the parity creation without much problem, I'd be very wary of restoring them if there was any kind of problem, and not much point in backing up if you can't restore

You state "other than WHS ones...". I looked around trying to find out if the actual PC backups can be mirrored, but most of the information I found was about backing up the WHS server and how it had been removed from PP1. I'm not sure if it was related to the PC backups as well.

Do you know if they can be mirrored?

If they are not, then I don't see an alternative but to use FlexRaid on the /Shared directories as well. Even if they change after every backup, if I do a rsynch every night, I should be able to recover in case of hard drive failure.

If they can be mirrored, I think I'm going to change my configuration to match your setup of using more mirroring. I like the system being able to handle the redundancy automatically without my involvement. I'm lazy like that

I have noticed that WHS doesn't move files around very much unless there are changes done to the file system. It took me a while to figure out that the reason my parity create set was failing, was due to having PC backup enabled and running during that time.

olgemap wrote:

Prelector wrote:

... Either way, I'd HIGHLY recommend you leave the "pc backups" from WHS out of any protection scheme you implement other than WHS ones... this is all database info, and as such, is very touchy about being played with. While you could probably run the parity creation without much problem, I'd be very wary of restoring them if there was any kind of problem, and not much point in backing up if you can't restore

You state "other than WHS ones...". I looked around trying to find out if the actual PC backups can be mirrored, but most of the information I found was about backing up the WHS server and how it had been removed from PP1. I'm not sure if it was related to the PC backups as well.

Do you know if they can be mirrored?

If they are not, then I don't see an alternative but to use FlexRaid on the /Shared directories as well. Even if they change after every backup, if I do a rsynch every night, I should be able to recover in case of hard drive failure.

If they can be mirrored, I think I'm going to change my configuration to match your setup of using more mirroring. I like the system being able to handle the redundancy automatically without my involvement. I'm lazy like that

I have noticed that WHS doesn't move files around very much unless there are changes done to the file system. It took me a while to figure out that the reason my parity create set was failing, was due to having PC backup enabled and running during that time.

Never underestimate the importance of the "laziness factor" in software development

That's my exact reasoning for my design also

I "thought" there was still a method for backing up the PC backups, in PP1... I'm not positive on that though, since I don't bother with it

I run backups every night, and the chances of losing a PC AND the WHS box both at the same time are pretty slim...

At a minimum, if you do decide to use FlexRAID for the pc backups, PLEASE set that up as a seperate backup scheme, with it's own datasets and parity set. This way, if you are forced to recover them, you KNOW to restore all the components, not just a particular drive. Remember, these backups are actually a living database on the WHS system, so restoring just one drive has a pretty good chance of really hosing some things up.

This message was edited 1 time. Last update was at 28/11/2008 23:26:18

Total n00b here, so bear with me, if you wouldn't mind. I'm still trying to grasp all this and need to go through the FlexRAID info several more times.

But, since this was posted several months ago, is everything posted above still relevant and accurate? I'm still building my server, but I'd like to play around with FlexRAID on the server as it currently stands with the hardware below, before I'm ready to implement it. Any suggestions for my plans below?

1. SuperMicro board with 6 SATA ports:
* 0 - 750GB WHS System drive
* 1 - 1TB Storage Pool drive
* 2 - 500GB Server Backup drive
2. SuperMicro SATA card with 8 ports:
* 0 - 1TB DVR recordings
* 1 - 1TB DVR recordings
3. Highpoint RAID card with 4 ports (none in use)

This is intended to be a monstrous home server with capacity for over 30 drives. My initial plans have been to use FlexRAID on the DVR recording drives (which will be added to over time), as well as for my photos, MP3s and DVDs. For the record, I'll be running SageTV with multiple HD-PVRs hooked up to DirecTV.

I planned to store important documents on the regular WHS drives, separate from FlexRAID. The only thing I'm currently using WHS for is nightly PC backups (2 computers, so far).

Would anyone care to take a stab at explaining the best way to set this all up on the server? Do I need to set up hard links (which I'm not at all familiar with, but will begin reading about), like the OP did? How should I add the drives to the server storage (i.e. outside the WHS storage pool somehow or as part of it)? Feel free to tell me to go read A, B, C, X, Y, and Z and/or that I'm completely headed down the wrong path.

Thanks in advance!

This message was edited 2 times. Last update was at 29/11/2008 01:32:41

1. Download LSE http://www.openegg.org/forums/posts/list/29.page
2. Create a folder in your C drive and name it whatever you want
3. For each drive/partition in your system, mount them into the folder you created

To reduce the need to resynchronize your RAID, you should setup the drives managed by FlexRAID outside of WHS's pool.
If you do that, however, you will not be taking advantage of WHS's disk free space management (WHS will not be moving the data around to balance the free space).

So, you can either:
- add all of your drives (except for the parity drives) to the WHS's pool and let FlexRAID RAID the whole pool
- or have a separate WHS pool (not touched by FlexRAID) and a set of drives outside of the pool (RAID'ed by FlexRAID)

In all cases, enable folder duplication of the folders that will contain your critical data.

You don't need to run LSE. There are already mappings done for WHS's hard drives on C drive that you can use. The mapping "C:\fs\<drive_letter>\DE\shares" are created on all the hard drives that WHS are storing data on. I didn't have any problems adding the shares even though some are hidden.

Unfortunately, the way WHS works is that it likes to fill a drive before it starts using another one, so all drive might not have the "shares" mapping on there initially. I ended up adding junk data (outside of WHS) to all hard drives to force WHS start using all hard drives. The other way is to start FlexRaid with the drives that WHS are using and then add a Parity source (DRU) to the exiting Parity set and do a rsynch. I haven't tried the last suggestion, but from reading here, I believe that should work.

Also, keep in mind that you have two directories created on the hard drives. One is "shares" and the other is "folders". The folders one is specifically for storing your PC backups and as suggested by Prelector, it's best to not include that in FlexRaid. The possibility of a PC needing a restore at the same time WHS has a hard drive failure is very slim.

I also use folder duplication by WHS for critical data since it makes is so easy to use. This is how my FlexRaid is setup:

Data:
1DRU
C:\fs\10\DE\shares
2DRU
C:\fs\15\DE\shares
3DRU
C:\fs\1D\DE\shares
4DRU
C:\fs\1J\DE\shares
5DRU
C:\fs\1L\DE\shares
6DRU
C:\fs\1N\DE\shares
7DRU
C:\fs\1R\DE\shares
8DRU
C:\fs\U\DE\shares

Parity:
1PPU
X:\

I find it useful to RAID my system backups.

I do that because:

I store more than one backup for most of my systems and find it useful to be able to restore a specific backup.
For instance, I have backups of my systems at these points: after fresh install-no drivers/after certain drivers were installed/after the apps were installed/after certain config changes.

Even if you store only one backup, taking the current system state might not be what you want since it will probably be "dirty".
If you lose your backup, you won't able to restore to that particular point you might have wanted.

Brahim wrote:To reduce the need to resynchronize your RAID, you should setup the drives managed by FlexRAID outside of WHS's pool.
If you do that, however, you will not be taking advantage of WHS's disk free space management (WHS will not be moving the data around to balance the free space).

So, you can either:
- add all of your drives (except for the parity drives) to the WHS's pool and let FlexRAID RAID the whole pool
- or have a separate WHS pool (not touched by FlexRAID) and a set of drives outside of the pool (RAID'ed by FlexRAID)

Does anyone know how frequently WHS moves data around? Hourly? Daily? I'm still VERY new to WHS.

What are the downsides to adding the drives to the WHS pool, other than having to resynchronize more often? It would seem that FlexRAID would be able to resynch pretty quickly, so the drives shouldn't be thrashing 24/7, right? Obviously, this depends on the answer to my previous question, which is why I asked.

Lastly (for now, that is), I keep seeing mention of just a single parity drive. I'm having a hard time believing that I could have 30 1TB hard drives with data and still only need a single 1TB drive for parity. This comes up because I'm building my server in a Lian-Li 383 server case which has 18 front 5.25" bays and I'm slowly adding 5-in-3 disk arrays from Addonics to ensure I'll always have room to grow. (I only have one 5-in-3 at the moment, though.)

Thanks, everyone. Looking forward to trying this out over the weekend.

This message was edited 1 time. Last update was at 04/12/2008 18:56:51

Yeah, it would be interesting to know how often WHS moves the data around.

As far as only needing one parity drive.... well, it is like magic.

But seriously, it is rather very simple.
The RAID4 engine does simple XOR.

The RAID6 and RAID∞ engines, however, are a totally different ball game.

skirge01 wrote:Lastly (for now, that is), I keep seeing mention of just a single parity drive. I'm having a hard time believing that I could have 30 1TB hard drives with data and still only need a single 1TB drive for parity.

There are I/O and performance limits to the number of drives you could realistically calculate parity for. But yes, only one parity drive is needed.

A *really* uneducated example:

Try writing an odd number of columns (your hard drives) and filling them with 0s and 1s for several rows. Now look at them and sum them up, saying that if there are more 0s the result is even and if there are more 1s the result is odd. Write that result in a new column (your parity drive).

Now imagine of taking out any single column (one hard drive failure). By looking at the parity column you can easily guess the value of the missing column for every row.

Obviously XOR operations don't work this way, so the odd number of columns (hard drives) is just to make the example work, you can have an even or odd number of hard drives, it makes no difference. But I believe the principle is close to this.

That's oversimplified, but I believe it can help in explaining parity (well, that's how I understand it, so it might as well be wrong. If it is, I'm deeply sorry

).

Forgive me if you guys have already answered these questions and I haven't "gotten it" yet. I'm trying, though!

I'm working to get this set up now and would like some clarification on what my brain has come up with:

I've decided to keep my data outside of the WHS data pool to cut down on WHS' balancing, but I still want to use the common shares like olgemap mentioned.

Q: Is this possible, or do those shares require that these all be WHS managed drives?
Q: If they do not need to be managed drives, how do I still add the drives to the WHS common shares?
Q: Building on the previous question, is this where LSE would come in?

I currently have 4 1TB drives outside of the drive pool, formatted as 64k clusters, with no path or drive letter assigned by WHS.

Q: I believe that the parity drive should use something like 4k clusters, right? But, that wasn't an option during the formatting process. Did I miss something there or do I need to format it in a special way in order to get 4k clusters?

When I go to create the array, I need to add/edit the "data=" and "parity=" lines to the create.txt file. I assume that I need to have the drives added to WHS (not necessarily the managed drive pool) in order to know what network path to specify in these lines.

Q: How do I know what paths to add to the create.txt file?
Q: Should I format the drives via Disk Management to add the drives with network paths, rather than assigning a drive letter?
Q: Once an array is created, is there any reason to run the create.txt again on the SAME array?

After the array is created, I run the rsynch.txt.

Q: Does anything need to be customized in there before I run it, such as where the flxr.meta file is stored?

After the array is set up properly, rsynch has been completed, and I decide to add another drive, I need to add the new path to the rsynch.txt file as a "data=" parameter. After the next rsynch.txt is run, I can either delete the new path from the rsynch.txt file or leave it there, since it will be ignored on subsequent runs. For good housekeeping, I guess it makes sense to delete it from there.

I'm eventually going to try out the WebGUI, but I wanted to make sure I understood what's going on with everything behind the pretty face.

I think that's it for now. Thanks in advance!

This message was edited 1 time. Last update was at 12/12/2008 16:47:15

skirge01 wrote:
Forgive me if you guys have already answered these questions and I haven't "gotten it" yet. I'm trying, though! I'm working to get this set up now and would like some clarification on what my brain has come up with:

I've decided to keep my data outside of the WHS data pool to cut down on WHS' balancing, but I still want to use the common shares like olgemap mentioned.

Q: Is this possible, or do those shares require that these all be WHS managed drives?
Q: If they do not need to be managed drives, how do I still add the drives to the WHS common shares?
Q: Building on the previous question, is this where LSE would come in?

First, you need to assign drives letters to your drives (use Remote Desktop to connect to WHS and then use Computer Management tools to assign the drive letters).
Once you have the drives letters assigned, you use LSE to map the drives to the common share folder.

I currently have 4 1TB drives outside of the drive pool, formatted as 64k clusters, with no path or drive letter assigned by WHS.

Q: I believe that the parity drive should use something like 4k clusters, right? But, that wasn't an option during the formatting process. Did I miss something there or do I need to format it in a special way in order to get 4k clusters?

FlexRAID imposes no restriction on how you format and what the cluster size of your drives should be.
Use whatever cluster is appropriate for your data.
Tests have shown that 4k works really well for NTFS volumes.

The FlexRAID buffer size, is something you need to play with to get the best performance out of it.
Only benchmarking will tell you which buffer size is optimal for your RAID configuration as it varies from system to system.
On my system, I use a 1MB buffer size and all of my drives are formatted with 4k clusters.

When I go to create the array, I need to add/edit the "data=" and "parity=" lines to the create.txt file. I assume that I need to have the drives added to WHS (not necessarily the managed drive pool) in order to know what network path to specify in these lines.

Q: How do I know what paths to add to the create.txt file?
Q: Should I format the drives via Disk Management to add the drives with network paths, rather than assigning a drive letter?
Q: Once an array is created, is there any reason to run the create.txt again on the SAME array?

After the array is created, I run the rsynch.txt.

Q: Does anything need to be customized in there before I run it, such as where the flxr.meta file is stored?

The 1.0 release of the CMD/Shell client will have the option of using a single configuration file for all your tasks.
http://www.openegg.org/forums/posts/list/74.page

For now though, you need to define a configuration for each task.
So, you need to edit the configuration file for each of the tasks you plan on using.
At the very least, you will need a create and and rsynch task configurations (so edit both).

The create task is executed only once.
The rsynch task is what you will use regularly.

Once you have your drive assigned with drive letter and mapped using LSE to the common share, you use common share location as paths for defining the data and parity properties.
Try reading Prelector's first post one more time.

After the array is set up properly, rsynch has been completed, and I decide to add another drive, I need to add the new path to the rsynch.txt file as a "data=" parameter. After the next rsynch.txt is run, I can either delete the new path from the rsynch.txt file or leave it there, since it will be ignored on subsequent runs. For good housekeeping, I guess it makes sense to delete it from there.

I'm eventually going to try out the WebGUI, but I wanted to make sure I understood what's going on with everything behind the pretty face.

I think that's it for now. Thanks in advance!

If you add a new drive and edit the data property of the rsynch config file, you might want to leave it there so you know of all the paths currently in your RAID.

Thanks for the quick reply. I'll check out LSE this weekend and see if I can get that working and I'll also re-read Prelector's post. Not that I'm going to get there anytime soon, but if I need to assign drive letters, wouldn't that mean I'm limited to "only" 26 drives? (I can't believe I just said that!)

This message was edited 1 time. Last update was at 13/12/2008 09:59:36

skirge01 wrote:Thanks for the quick reply. I'll check out LSE this weekend and see if I can get that working and I'll also re-read Prelector's post. Not that I'm going to get there anytime soon, but if I need to assign drive letters, wouldn't that mean I'm limited to "only" 26 drives? (I can't believe I just said that!)

Ah, nope you are not limited to 26 drives.

Read this: http://ask-leo.com/26_drives_is_there_a_way_around_the_26_drive_limit_in_windows.html

I guess I should stress more often that you don't really need LSE at all.
You can create the hard links as explained in that article or this other way: http://articles.techrepublic.com.com/5100-10878_11-6177180.html

This message was edited 1 time. Last update was at 12/12/2008 23:03:21

Great info there! Thanks for that first link. It really helped to start making sense of these shares in WHS.

Second one, however, is invalid.

I'll do some searching on that site and see if I can find the article you were trying to send me to.

EDIT: Nevermind, I just noticed the period after the link.