Re: btrfs raid10 rebalance questions

[DanglingPointer <danglingpointerexception@gmail.com>]

Hi Todor,

Have you tried looking at the new documentation? 
https://btrfs.readthedocs.io/en/latest/

Could someone respond to Todor?  Many have similar questions and 
experiences.  Thanks in advance!


On 22/5/23 23:02, Todor Ivanov wrote:
>       Hello,
>
>       We have a debian10 system with 6x16TB in btrfs raid10. In the
> past we hit an issue with lack of space, which we resolved with data
> rebalance, but some questions left unanswered:
>
> https://unix.stackexchange.com/questions/743528/btrfs-snapshot-fails-with-no-space-left
>
> We will be very happy if you can answer or give guidelines for at
> least the following:
>
> - How often should we run btrfs balance? Trying to use some logic and
> looking at https://docs.nvidia.com/networking-ethernet-software/knowledge-base/Configuration-and-Usage/Storage/When-to-Rebalance-BTRFS-Partitions/
> looks like a good example, but this is not for RAID10. How do we
> calculate chunk size correctly and should we alter Device Size beause
> of data duplication?
> - Is it dangerous and should we rebalance metadata as well, having in
> mind we use btrfs-progs v4.20.1, kernel 4.19.0-16-amd64 and btrfs
> raid10? What is an optimal value for musage?
> - What does it mean when "btrfs fi us" is showing a lot of
> "Unallocated" space, and yet we ran into the out of space issue
> (probably on Metadata data - subvolume snapshot), why isn't Metadata
> expanding into that Unallocated space automatically?
>
>
> Kind regards,
> Todor