Well I have been using Plex neigh on 4 to 5 years now.
As one would expect, my drive farm has gotten very large.
My question is, what issues, concerns, or considerations do I need to worry about implementing Drive Pool after the fact.
What will the PMS ramifications be?
Thanks for any advice or tips.
Expect very few issues or special needs EXCEPT - NEVER place “The path where local application data is stored” in the pool. It will NOT work. I use DrivePool and have no issues with it.
Having said the above I have found that on my system everything works better with a computer running DrivePool and sharing the DrivePool drive on the network separate from the computer running Plex. While it worked OK to have both Plex and DrivePool on the same computer it seem much easier to maintain with the two computers separate.
One more thing I would turn off “Empty trash after scan” just in case there is a glitch in the pool and I would set all the pooled drives so they never sleep or spin down.
It is well to remember that the duplication feature in DrivePool works well BUT it is not a backup but rather it creates solid redundancy. For backup you really need separate drives (or other storage) outside the Pool.
@Elijah_Baley
So the transfer to using Drive Pool will be transparent, sort of. Right?
All I should have to do is point libraries at the new Drive Pool drive??
I will be installing DP on the PMS box. What issues are there with that??
With DP setup, will Plex still see the discrete HDDs while DP is installing and or doing it’s thing.
In other words, I will be able to access all my HDDs with a single drive designation but will I still be able to access them directly/individually??
Thanks for the info.
@jjrjr1 said:
@Elijah_Baley
So the transfer to using Drive Pool will be transparent, sort of. Right?
All I should have to do is point libraries at the new Drive Pool drive??I will be installing DP on the PMS box. What issues are there with that??
With DP setup, will Plex still see the discrete HDDs while DP is installing and or doing it’s thing.
In other words, I will be able to access all my HDDs with a single drive designation but will I still be able to access them directly/individually??
Thanks for the info.
To answer your last questions first, yes, but with a few caveats.
Plex can still see all the drive just as it does now, again with a few caveats.
The only two issues I ran into with DP and PMS on the same computer was the computer was slow to startup sometimes and sometimes the DrivePool service did not start requiring a reboot to get it going. Those could have bee mainly because the computer was probably a but under-powered for the combined tasks.
Now to address the first parts and the caveats I mentioned.
When you install DrivePool it creates a hidden sub-directory on each drive you put it the pool. Those sub-directories are linked and form the pooled drive which is given a new drive letter. Any content that is not in one of the hidded subdirectory on one of the pooled drives is not in the pool and is not managed by DrivePool.
The best way, or at least the fastest, to get your content into the pool is to go onto each drive and move the content into that hidden sub-directory on that drive. Directly moving the content like that is fast but moving content from a drive to the DrivePool drive goes through the system just like moving from one drive to another so it is much slower. Once you do that the content shows up in the Pool but it still can be directly accessed as well by simply going into that hidden sub-directory but, once the content is in the pool accessing it through the DrivePool drive is the best way.
The DrivePool drive behaves exactly like any other windows drive and can even be shares on the network and is seen by Plex as just another drive.
One more suggestion is that you install the “Load Balance” plugin and duplicate your media folders as desired (If you ever change the name of a duplicated folder you will have to set the duplication on it again) and then let DrivePool do all the work. Do not try to micromanage DrivePool, just let it work for you.
DrivePool does protect you very well from a single drive failure but it does not protect from your own deletions so be kind of careful as you move stuff around.
I hope the above helps some.
BTW: StableBit’s forums are pretty good and my experience that any requests for help get answered within 24 hours, usually less, by one of the developers.
Edit: One more thing - DrivePool never will erase your content during install or setup.
@Elijah_Baley
Thanks so much.
Very helpful!!!
@Elijah_Baley
Man you are right.
StableBit support is damn good.
I submitted some technical questions this morning (Saturday) I have already received an answer.
A New experience after dealing with this pathetic support system.
lol.
@jjrjr1 said:
@Elijah_Baley
Man you are right.
StableBit support is damn good.
I submitted some technical questions this morning (Saturday) I have already received an answer.A New experience after dealing with this pathetic support system.
lol.
I saw your questions over in their forums and the answers you got are clearly complete and accurate but I wanted to give you a simple answer to one of them:
Lets suppose you have two drives both with directories AA and BB and CC the drive have letters assigned to them of F: and G:
On drive F: the directories contain:
AA
… File 1
… File 2
BB
… File 3
… File 4
CC
… File 5
… File 6
On drive G: the directories contain:
AA
… File 7
… File 8
BB
… File 9
… File 10
CC
… File 11
… File 12
You then create a Pool with both drives and that adds a directory called PoolPart12345 to drive F: and a directory called PoolPart67890 to drive G:
Also created at that time is a new virtual drive say M: that is really the combination of the contents of both PoolPart directories on the two real drives. That drive starts out empty.
Now to get content on that drive you do it the easy way and go to drive F: and move (Not copy) the directories AA, BB and CC to the PoolPart12345. Note that is exactly the same process as copying a regular directory into another directory on that same drive.
After that process if you go to M: you will see:
AA
… File 1
… File 2
BB
… File 3
… File 4
CC
… File 5
… File 6
Then you go to drive G: and move AA, BB and CC into PoolPart67890. You may get the Windows message about combining directories and you should answer “Yes” to that question.
Once you do that and go to drive M: you will see:
AA
… File 1
… File 2
… File 7
… File 8
BB
… File 3
… File 4
… File 9
… File 10
CC
… File 5
… File 6
… File 11
… File 12
It may take a little while for the moved files to show up on M: but it should be just a few seconds no longer than about a minute.
That is simply stated that if two or more PoolPart directories that are pooled contain the same directory name then the content of the like named directories will get combined in the pool.
I hope that clarifies the good technical info you received from StableBit and I hope it was not too unclear. 
Only just seen this thread. I have been using Drivepool since the very earliest alpha.
A year or two longer than I have been using Plex.
I have had zero issues in all that time. Unlike @Elijah_Baley I have also had no problems running Drivepool on the same server as Plex.
For example I haven’t encountered the pool not starting up.
One thing that may help if you ever get to the ten or more hard drives (as I currently am) is to create network shares for your libraries and point Plex to those. I’m guessing this is what Elijah does if he is running the server and pool seperately. But it just feels snappier even if running on the same device.
As for the support in Drivepool well you both said it all.
For a project with a lone developer and Drashna who deals with support issues it’s a truly incredible product.
It isn’t just Plex who could learn a thing or two about customer support from them.
@Elijah_Baley
Thanks for the additional info.
Question
After all the drives are seeded, can files be copied directly to M: (Per your example)
Wish me luck… Gonna try the implementation this week…
LOL
@jjrjr1 said:
@Elijah_Baley
Thanks for the additional info.
Question
After all the drives are seeded, can files be copied directly to M: (Per your example)Wish me luck… Gonna try the implementation this week…
LOL
Of course the content can be copied/moved pretty much as you wish. Again the easiest and fastest way is as I mentioned in my earlier post by moving the files or directories directly into the PoolPart directory. BTW that PoolPart directory is hidden (but otherwise it is a normal directory) so whatever tool you use will need to be configured to allow viewing of hidden directories.
Once the content is moved into the various PoolPart directories it is by far the best way to manipulate the content is through the pool drive. (In my example drive M:)
DrivePool does not compress or use any form of “parity” drives or anything else like a Raid type system. All files are kept it their native format and can be read directly by Windows. That means that any duplication takes exactly the same space as the original but drive are cheap enough that I find keeping two copies of all important files to be the best way for redundancy.
I once had a computer with DrivePool failed (The mother board failed for some reason) and all I needed to do was install DrivePool on the new computer and move my license to a new computer (Which process the good folks at DrivePool made painless) and attach all my drives to it and DrivePool recognized the pool and everything was there just as before. No fuss no muss.
I then pointed my Plex libraries (I run Plex on a different computer) to the Pooled directories on the new computer and Plex was running correctly again after a library scan to detect the new location.
Oh, one more time, do NOT place “The path where local application data is stored” on the pooled drive. It simply will not work correctly and your database will get corrupted. Keep that directory on a drive that is not pooled at all is best practice or outside the PoolPart directory on a drive that is pooled but that is a little risky because the data in the pool on that drive could grow to the point where the Plex data cannot grow as it should. I always try to have one large drive (1tb or better) in my Plex computer that holds Plex’s data. That drive is usually my boot drive as well and my Plex computer generally has no other tasks.
Edit: It occurs to me that you were possibly really asking if the pooled drive (M:) could be treated as a regular drive after the initial content was moved to the various PoolPart directories and the answer is YES, You computer will not really know that the new drive is not a regular hard drive. It can be written to and gave files deleted from just like it was a very large drive attached to your computer.
Well I have started the move to Drive Pool.
Gees… a big job. LOL
I recommend you install DP before you get a huge library…
@jjrjr1 said:
Well I have started the move to Drive Pool.
Gees… a big job. LOLI recommend you install DP before you get a huge library…
In what way? If you are simply adding your existing data into the pool it takes about 10 seconds. You probably want to check Elijahs replies because I’m fairly sure he explains it.
40TB of data can be added to the pool in about 2 minutes.
@HitsVille said:
@jjrjr1 said:
Well I have started the move to Drive Pool.
Gees… a big job. LOLI recommend you install DP before you get a huge library…
In what way? If you are simply adding your existing data into the pool it takes about 10 seconds. You probably want to check Elijahs replies because I’m fairly sure he explains it.
40TB of data can be added to the pool in about 2 minutes.
Yep, the only problem comes in if you are creating the pool on drives that do not already contain your data. Although when I first setup DrivePool I only had about 10tb of data and it took me more like 20 minutes instead of 2.
Part of that I believe that I was optimizing my directory structure at the same time so I spent time changing directory names and moving stuff around. BUT the actual data import was very very fast. As long as you use move and not copy and as long as you moves are all on the same drive there is almost no time spent moving data to the pool.
Well yeah but with regards to moving to drives that don’t contain the data already surely that simply applies to every O/S and it certainly nothing to do with Drivepool.
And of course if you are setting up with a different structure it will obviously take longer…but even then as you mention… only a matter of a few minutes longer.
I’m pretty sure @jjrjr1 is doing it all through the Drivepool interface despite you correctly giving him the better way of achieving it.
Lastly if any micro-management is then required it can be done in the background by Drivepool over a period of time.
Well
I guess I might be doing it wrong but I do not know how…
I am moving the files from the original drives into the pool, The files actually move and take the same time as if I were to copy to any other drive.
I am not doing it thru the Drive Pool interface, Just Windows Cut and Paste moves.
The other challenge has been about 10% of the files moved make new DB entries and I have to go and Fix Match.
Just a bigger job than I thought.
I would have thought DP would not have copied a file somewhere but just made the directory entry for the new location if the file was already on a drive in the Pool.
IT does not seem to work that way. But I prolly am doing something wrong…
You should be moving the files into the PoolPartxxxxx directory from the same drive not into the Pool drive itself.
That is an example: move H:\Movies to H:\PoolPartxxxxxx
Remember that the PoolPart directories are hidden so you have to set your file viewer to display hidden files to make it work.
I can see those directories. (Show Hidden Set)
I must have missed that in the documentation…
Can I do that even if I already started doing a file move at the pool level??
@Elijah_Baley
That Works!!!
Thanks
@jjrjr1 said:
I must have missed that in the documentation…
Yeah i dont think its actually in the Drivepool documentation because it is kind of a hacky way of doing it.
However it was the one of the Stablebit guys who first mentioned it in their forums.
Hopefully it should save you some time.
Since you are only now starting to move files into your pool, here is something critical you need to know:
By default DrivePool spreads all data supplied into the pool across all drives. Which means if one hard drive dies, you have no idea what files are missing. This is a huge problem.
The default configuration only makes sense if you want full duplication by DrivePool -> 50% usable capacity. Since you are using Plex, most storage goes to movies/series. Because of that I suggest the following:
Configurable Folder Split Levels would be an more ideal solution but the sole DrivePool developer “Alex” personally doesn’t see the need for it -> low on priority list and will most likely never happen.
