XFS Developer Takes Shots At Btrfs, EXT4

[jwilliams]

This is not true anymore.

https://fedoraproject.org/wiki/Features/F17Ext4Above16T

Good to know. Unfortunately, from your link:

"e2fsprogs-1.42 has already been released with initial support greater than 16T, and kernel code already has support as well. There are rough edges to work out, and there is testing to be done."

So, I would not trust it for at least 6 months, if not a year, until they have worked out the "rough edges" and tested it thoroughly.

[RahulSundaram]

Good to know. Unfortunately, from your link:

"e2fsprogs-1.42 has already been released with initial support greater than 16T, and kernel code already has support as well. There are rough edges to work out, and there is testing to be done."

So, I would not trust it for at least 6 months, if not a year, until they have worked out the "rough edges" and tested it thoroughly.

If you have over 16 TB drives that you handle regularly, then you are a good candidate to participate in testing and report any issues. It won't happen automatically.

[jwilliams]

If you have over 16 TB drives that you handle regularly, then you are a good candidate to participate in testing and report any issues. It won't happen automatically.

And I won't test it with important data when XFS works fine for me already.

[frantaylor]

And I won't test it with important data when XFS works fine for me already.

Given that any filesystem testing should be done off-line, with backed up data, I wonder how you can make the assertion that testing will endanger the integrity of your "important" data.

[jwilliams]

Given that any filesystem testing should be done off-line, with backed up data, I wonder how you can make the assertion that testing will endanger the integrity of your "important" data.

It won't because I will not be testing it! Geez, some people...

[RahulSundaram]

And I won't test it with important data when XFS works fine for me already.

If XFS works fine for you, then the support of Ext4 for it shouldn't matter to you at all. I thought since you bought it up you have a interest in it.

[droidhacker]

Of course, for mobile and desktop Linux users this problem is largely mute.

That should be ***MOOT***, not "mute".

[jwilliams]

That should be ***MOOT***, not "mute".

Unless he meant that mobile and desktop linux users were having trouble muting their sound.

[houkouonchi]

My mind boggles that ext4 filesystems STILL are not supported with ext-tools for filesystems larger than 16TB.

XFS is the only filesystem I trust to use for >16TB. Other possibilities are JFS, ZFS, and btrfs, but none of those are mature, stable, and actively developed on linux.

Maybe one day btrfs will be stable and not have any flaws (very slow performance in certain situations). But that day is far from today.

I agree with its unbelievable that ext4 does not have *proper* > 16 TiB support. That being said I think I would disagree with JFS and ZFS not being stable and mature.

This. Even though JFS was designed for "huge" FS too, I understand it was only run on such with AIX, and so the linux jfs tools could not cope with >16tb until somewhat recently.

Not true. There was a bug in mkfs which had problems with > 32 TiB file-systems (not 16). This was fixed now and I have a file-system over 64 TiB running JFS:

root@dekabutsu: 12:42 AM :~# df -H
Filesystem             Size   Used  Avail Use% Mounted on
rootfs                 129G   103G    27G  80% /
/dev/root              129G   103G    27G  80% /
udev                    11M   238k    11M   3% /dev
/dev/sda1              129G    78G    52G  61% /winxp
/dev/sdd1               36T    26T    11T  72% /data
/dev/sde1               84T   5.6T    79T   7% /data2
tmpfs                   13G   144k    13G   1% /dev/shm

It might be getting time to switch to XFS as it has imrpoved a lot over the years but I had pretty major issues with corruption (especially on crash/powercycle, kernel panics, etc..) in the past. Granted this was in the earlier 2000s but once you have a bad experience with a fs that caused dataloss its hard to go back.