Can I cancel BTRFS balance on pool raid1 without repercussions?


Recommended Posts

A few here have been helping me with disk configuration, etc... with my new netapp ds4486. I settled on using the pool feature for the disks and splitting the 48 disks into 2 pools. I had already created a pool with about 12 disks to try and then added in the additional 12 after I had added 10tb of data or so to the pool. Now it is taking FOREVER to "balance" the raid1 with the new disks added. I realized that I actually want to do raid5 to increase the available capacity to me to the max and not be able to lose the entire array with a single disk failure. Ideally, raid0 would be best to give me the most space, but it seems I would lose the entire array if a single disk fails, which I don't want. I don't mind if I lose the data on that disk, but I don't want to lose the entire array/pool of data at once from one disk failure.

 

To the point. Can I cancel the balance and then immediately switch it to raid5? The balance is at 43% done and it has been about 2 days?

Edited by live4soccer7
Link to comment
56 minutes ago, JorgeB said:

Also I would recommend using smaller pools, especially if it's large capacity disks, much easier to deal with, or any operation needed can take days or even weeks, they can still be all using the same share.

They are 4TB disks. The idea is to maximize space. Each raid 5 I create, I would lose 4tb worth of space.

 

So, when I go to the pool for the raid and click to change to raid5, it says *see help. I don't see this and there is no additional information in the wiki/manual.

 

https://wiki.unraid.net/Manual/Storage_Management#Change_Pool_RAID_Levels

 

In the manual it says: BTRFS supports raid0, raid1, raid10, raid5, and raid6 (but see the section below about raid5/6)

I can not find this section. 

Link to comment
41 minutes ago, JorgeB said:

If this is for farming chia that's what I would recommend, do multiple raid0 pools with 4 or 5 disks max, they can still be all on the same share and if you lose a disk you only lose a small pool.

 

Great idea! I just did raid 5, but it is giving me some wonky numbers for free/available space. I'm not sure if it is a bug within unraid or some other issue. I have two identical raid5 setups and one says 92tb available and the other is 96. Regardless, I'm going with what you suggested on the raid0. I just did half the disks with 4 raid0 pools (6 disks a pool to keep numbers even).

 

When I create a new share, how can I add the "pools" to the share? I just see use cache and then I can select one pool/cache, but not multiple.

Link to comment
6 minutes ago, live4soccer7 said:

When I create a new share, how can I add the "pools" to the share? I just see use cache and then I can select one pool/cache, but not multiple.

User shares are not (yet) prepared to work with multiple pools, but since this is a write once situation it's easy to accomplish, create the pools share as use cache=only and select the first pool, once that pool is full go to the share settings and select to use the next one, and so on, chia just needs to be mapped to /user/plots and it will access all the pools that have that share.

Link to comment
10 minutes ago, JorgeB said:

User shares are not (yet) prepared to work with multiple pools, but since this is a write once situation it's easy to accomplish, create the pools share as use cache=only and select the first pool, once that pool is full go to the share settings and select to use the next one, and so on, chia just needs to be mapped to /user/plots and it will access all the pools that have that share.

 

Easy enough, but once you change the pool that is used as the "cache" in the share settings then there is only one pool that has that share unless this persists in the DB for past selections???

Link to comment
1 minute ago, live4soccer7 said:

there is only one pool that has that share unless this persists in the DB for past selections???

There's only one pool for any new written data, if you access the share you'll still see and access the data form any pool (and/or array disk) that contains that share.

Link to comment

Interesting. When I did a mv command in mc it didn't actually put the file on the "cache drive" pool, but just instantaneously changed the share the file belonged too. Is this expected? I was moving from a share on the actual unraid array.

 

I have "Use Cache Pool" set to ONLY

Edited by live4soccer7
Link to comment
9 minutes ago, live4soccer7 said:

When I did a mv command in mc it didn't actually put the file on the "cache drive" pool,

It will if it's being moved from a different share and you are using user shares (/mnt/user/share) as source and dest, never move from user shares to disk shares or vice versa, fine to use disk shares only.

Link to comment
4 minutes ago, JorgeB said:

It will if it's being moved from a different share and you are using user shares (/mnt/user/share) as source and dest, never move from user shares to disk shares or vice versa, fine to use disk shares only.

Yes, I was going from /mnt/user/share1 --> /mnt/user/share2 (raid0 pool)

Edited by live4soccer7
Link to comment
8 minutes ago, JorgeB said:

Then it should go the the pool, if cache for share2 is set to "only"

Weird. maybe the system needs a reboot or something because that's definitely not happening.

 

Could the "Copy on Write" setting have anything to do with it? It is set to auto.

Edited by live4soccer7
Link to comment
1 minute ago, live4soccer7 said:

Weird. maybe the system needs a reboot or something because that's definitely not happening.

I think that is expected behaviour when you use mv on a User Share.   The Linux implementation of mv first tries a rename and only if that fails does a copy/delete.   In this case that works so the file remains on the drive (Linux does not understand User Shares).  You need to do an explicit copy/delete to get the file onto another dtive.

Link to comment
1 hour ago, live4soccer7 said:

Thanks. That's kind of what I was beginning to think. When I hit "F5" nothing comes up to actually copy, as if the button doesn't actually work. I'm a little new to MC and started recently using it over Krusader

you should be able to ALT+number.

For example ALT+5 (from the top row above letters) equals F5.

Link to comment

Well.... I learned the hard way that if you don't do background and your ssh loses connection then the transfer will fail. That lost about 16 hours. haha.

 

Is there a way to check progress of the background process? I started a new one. Also, what is the behavior of "background" for copying a file if it already exists in the destination? I see that it doesn't ask you or mention it. I had too many files to manually go through, so I just copied all the original files to the destination again.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.