[SOLVED]Upgrade xfsprogs to 3.2 and make new checksum option default


Recommended Posts

 

 

What I understood reading xfs docs:

Kernels above 3.14 support v5 of xfs without any additional work,

To use v5 you have to do mkfs.xfs with additional switches and that will create additional  structures on disk allowing new functionalities,

UnRAID will not have problems handling either old or v5 format as it has kernel new enough to support both,

Newest version of xfsprogs is needed to format HDD with v5,

Upcoming xfsprogs 3.3 will format disks with v5 by default (no cmdline switches needed).

 

Seems pretty straightforward to me and no harm is done to anybody.

 

Ok, I misunderstood your post. So if someone is formatted today with xfs and we include new progs and they want to use xfs v5, you are saying they will have to reformat, correct?  But if they don't care about v5 and we update the progs, even to 3.3 when released, it won't force them to reformat.

Link to comment

 

 

What I understood reading xfs docs:

Kernels above 3.14 support v5 of xfs without any additional work,

To use v5 you have to do mkfs.xfs with additional switches and that will create additional  structures on disk allowing new functionalities,

UnRAID will not have problems handling either old or v5 format as it has kernel new enough to support both,

Newest version of xfsprogs is needed to format HDD with v5,

Upcoming xfsprogs 3.3 will format disks with v5 by default (no cmdline switches needed).

 

Seems pretty straightforward to me and no harm is done to anybody.

 

Ok, I misunderstood your post. So if someone is formatted today with xfs and we include new progs and they want to use xfs v5, you are saying they will have to reformat, correct?  But if they don't care about v5 and we update the progs, even to 3.3 when released, it won't force them to reformat.

 

Yes Sir, you got it right.

 

Link to comment

Well that's a bummer. I literally just formatted my drives to XFS last week. Although it doesn't sound like the new features are must haves.

 

They are not. 

 

...It hardens the filesystem against power interruptions...Recovery from a power loss is the one weakness XFS seems to have over other traditional filesystems.

 

I have not seen any issues with XFS due to power interruptions in my testing.  I have "pulled the plug" many a time on test systems to see what happens and I have yet to see any data loss, file corruption, or any other issues with XFS.  I'm curious where this information about XFS and power loss is coming from?  Online research or direct testing?

Link to comment

...It hardens the filesystem against power interruptions...Recovery from a power loss is the one weakness XFS seems to have over other traditional filesystems.

 

I have not seen any issues with XFS due to power interruptions in my testing.  I have "pulled the plug" many a time on test systems to see what happens and I have yet to see any data loss, file corruption, or any other issues with XFS.  I'm curious where this information about XFS and power loss is coming from?  Online research or direct testing?

 

It has to be understood that issues will only manifest if you happen to be writing to the disk when power loss occurs, or if the sync is delayed for any reason.

 

It's an issue with two sides - on one side much of the reputation XFS has probably stems from a problem that was fixed back in 2007. Users of the filesystem prior to that fix experienced files that looked intact and had the right filesize, but contained zeroes (or garbage) instead of the intended content. Users who encountered that problem seem to have been scarred for life and continue to color much of the commentary that can be seen today on the internet.

 

On the other side, the XFS project team does say the v5 format is meant to bring reliability enhancements, so they must have felt improvements were warranted. See the xfsprogs 3.2 announcement: http://article.gmane.org/gmane.comp.file-systems.xfs.general/62552

 

It's hard to think of another scenario than power losses (or flaky hardware) where these enhancements could bring enhanced reliability, and it will obviously have required a significant development and testing effort to implement, so it would seem logical to conclude it was called for.

 

Jon, I am curious to know - did your testing include XFS-formatted cachedrives at all? Or powerloss while mover is running?

 

Link to comment

Are you considering it for next beta/rc?

If you are hesitant could you provide it as a plugin or just binaries for those willing to test it?

 

We're still working on getting it to compile.  There's something funky with this new version that's just being a little more troublesome.

Link to comment

xfsprogs 3.2.2 has been added to the next release (and added

-m crc=1,finobt=1

to mkfs.xfs).  Bear in mind, xfs CRC only applies to metadata, whereas with btrfs, CRC is calculated over all 'extents' whether used for data or metadata.  This is a "half solution" IMHO and why our interest in this feature has been basically a *yawn*.

Link to comment
  • 3 months later...

Hello,

 

I completely missed all the progress that have been made on this topic. I'm on 6.0.1 so it looks like I have the xfsprogs 3.2.2. ;D ;D ;D

 

My last question before I do the big jump is: Are the additional flags (-m crc=1,finobt=1) used by default when I click on the Format a new drive button using XFS filesystem?

 

Thank again for the effort.

Alphazo

Link to comment

Hello,

 

I completely missed all the progress that have been made on this topic. I'm on 6.0.1 so it looks like I have the xfsprogs 3.2.2. ;D ;D ;D

 

My last question before I do the big jump is: Are the additional flags (-m crc=1,finobt=1) used by default when I click on the Format a new drive button using XFS filesystem?

 

Thank again for the effort.

Alphazo

 

Yes.

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.