a duplication question.

[Toyebox]

Hey all (dev team and others!), 

 

Thanks in advance for any help you may provide! First off, I am a long time hardware raid user.. This is the first time i have ever migrated to any sort of software based storage pools. So far i love it!  Have a couple questions i couldn't find though. Maybe yall can help me!

 

So i currently have a small pool with five 2TB drives with duplication enabled (x2). I am trying to figure out what would happen if i move up close to the 9TB pool storage cap and i lose a drive. Let me rephrase... Lets say I have a failing drive and i want to remove it.. does it take all the current files and move them to the other hard drives? What if there is not enough space? On that note.. Is the duplication process a parity type with multiple pieces spread among all drives?  Sorry for the jumble of questions.. Just trying to figure what will happen when i come close to my pool limit and i get a drive failure is all.

 

Thanks!!

 

Matthew

[Christopher (Drashna)]

Sorry for the delay in getting to you!

And I'm glad to hear that you're enjoying the software so far!

 

Files in the pool are stored as the actual files on the pooled disks in a hidden folder.  We don't break the files up at all, and there is no parity used. It's "1:1 file based mirroring" (meaning there are two copies of the duplicated file on two different disk).

 

So if a disk does fail, the remaining contents of the pool remains intact, without any problems. Any unduplicated data may be lost, but any duplicated data is accessible, and will be re-duplicated once the missing disk is removed.

 

When you remove a drive from the pool, yes, it moves the files from the one disk and moves it to the other drives in the pool. Normally.  This can take hours depending on how much data you have, the number of files to be moved and other factors. 

 

I say "normally" here, because there are a couple of options for disk removal. Namely, "Forced damaged disk removal", which skips files that are not properly readable, and leaves them on the disk; and "Duplicate data later", which skips the duplicated data, removes the disk and then re-duplicates the data. 

This second one sounds like it may be what you want to do here. If you thing (or are sure) the drive is failing, it will help to quickly remove the drive (if it's filled mostly ore completely with duplicated data).

 

And if you don't have enough space, it will error out. Either during the removal process, or later on when it's re-duplicating data. But this depends on the way the drive was removed from the pool.

At this point, you'd want to replace the drive with a new one, so it has space to re-duplicate properly.

 

 

Sorry for the shameless self-promotion here, but ....

 

And since this is relevant to ... well, drive failure.... StableBit Scanner works very well with StableBit DrivePool as well. You may want to check it out. Not only does it monitor the SMART attributes of the drives (which can give pre-failure warnings), but it does a surface scan of the drive which will let you know if there are problems reading the disk, and can potentially trigger the drive's built in error correction to fix bad sections on the disk before they become an issue (a process usually referred to as called "data scrubbing"). 

 

Additionally, if you have both StableBit Scanner and StableBit DrivePool installed on the same system, DrivePool will grab information from Scanner. And if Scanner detects damage on the disk, DrivePool will automatically attempt to move all of the data off of the disk to help prevent data loss due to corruption/damage.