Ubuntu 9.04 Receives EXT4 Support

[phoronix]

Phoronix: Ubuntu 9.04 Receives EXT4 Support

With the EXT4 file-system having been stabilized with the Linux 2.6.28 kernel, the Ubuntu developers are preparing to adopt this evolutionary Linux file-system update. EXT4 will not replace EXT3 as the default file-system until at least Ubuntu 9.10, but as of yesterday, Ubuntu 9.04 now has install-time support for EXT4. In this article we are looking at the EXT4 support within Ubuntu as well as providing a few Linux file-system benchmarks from a netbook-embedded solid-state drive. In this article we have published Ubuntu benchmarks of EXT4, EXT3, XFS, JFS, and ReiserFS file-systems.

http://www.phoronix.com/vr.php?view=13363

[Tares]

Yay, I can switch / to ext4 with new release ^^;

[d2kx]

So what year does Ubuntu get the new theme which should get into 8.04?

@Topic: Great, will check it out in a few minutes, the Kubuntu-daily installer crashes for me for some time now :)

[M1AU]

Hopefully it'll be stable enough as I'm also very interested in switching to ext4 for some reason.

[val-gaav]

Hopefully it'll be stable enough as I'm also very interested in switching to ext4 for some reason.
Me too though I know why ...

Long disc checks are killing me, so 10 time speed up sounds very nice to me ... I wonder why those are needed, when I used RaiserFS there were no such checks there.

[Louise]

What I would really like to know is what the uncertainties are on those plots.

1% 5% 10%?

Results without error bars are worth nothing, as it is impossible to tell if all we see is noise.

[Amit]

Were these tests done on the hardware with a clean install each time with each file system as big as the others and occupying the same disk geometry?

[Michael]

Were these tests done on the hardware with a clean install each time with each file system as big as the others and occupying the same disk geometry?

Yes, we always do fresh installs each time.

[ashayh]

Please conduct a test with the latest version of btrfs.

Thanks

[Yfrwlf]

Please conduct a test with the latest version of btrfs.

Thanks

I thought BetterFS wasn't even fully implemented yet, or can it do most all the basics already? Regardless, I'm sure it won't be nearly as fast as later on when everything gets stabilized more, but benchmarks of it would indeed be interesting.

And don't forget Tux3! ^^

As far as EXT4 is concerned, it looks great, thanks for the benchmarks!

[TRS-80]

I read somewhere that MySQL performance was much improved with ext4 - does PTS have any MySQL or PostgreSQL benchmarks?

[crispy]

This test will probably affect what fs I will format my newly cleared 250 gb disk to. Although speed of the fs is a concern I am also concerned about being able to put the drive to sleep. Currently I have mounted my ext3 disks as ext2 so the journal wont be written every 5 sec which makes it impossible to put them to sleep with hdparm -Y /dev/sdX :(

[laurencevde]

Could you please test random write performance? The tested ssd has horrible (re)write-latencies, and filesystems can have a significant impact on that(the more small writes, the worse).
Mine's currently running udf...

[kgonzales]

Ext4 was released in Fedora 10, and will also be released as a tech preview with Red Hat Enterprise Linux 5.3.

You act like Ubuntu is trailblazing here.

[sc3252]

Ext4 was released in Fedora 10, and will also be released as a tech preview with Red Hat Enterprise Linux 5.3.

You act like Ubuntu is trailblazing here.

That is total BS, he did benchmarks before with fedora 9, and even mentioned that fedora had it in their distro well before ubuntu did.

"While the EXT4 file-system was marked as "ext4dev" within the kernel to signify its development status, Red Hat has been shipping EXT4 as an install-time option since Fedora 9. This support has continued into Fedora 10 with the newest kernel components. When installing Fedora, to create an EXT4 file-system you must append "ext4" to the boot line when starting the Anaconda installer. Beyond the Fedora option, EXT4 has yet to find itself as an option in many distribution installers."
http://www.phoronix.com/scan.php?page=article&item=ext4_benchmarks&num=1

[WSmart]

I'm excited about Solaris ZFS. Wonder how that would stack up.

Surprising how poorly EXT3 did.

[bugmenot]

i question the accuracy of benchmarks with these OCZ drives considering problems documented at http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3403&p=8

[clickwir]

Where can I download the ISO that was used here?
(Preferably Kubuntu)

Right now I'm using Alpha 2 alternate installer and it doesn't have ext4.

[d2kx]

Where can I download the ISO that was used here?
(Preferably Kubuntu)

Right now I'm using Alpha 2 alternate installer and it doesn't have ext4.

http://cdimage.ubuntu.com/kubuntu/daily-live/current/

You could also wait for Alpha 3 which is arriving tomorrow, but tere shouldn't be any changes to this one really.

[corepl]

These options could noticeably affect performance.

From Documentation/filesystems/ext4.txt


- When comparing performance with other filesystems, remember that
ext3/4 by default offers higher data integrity guarantees than most.
So when comparing with a metadata-only journalling filesystem, such
as ext3, use `mount -o data=writeback'. And you might as well use
`mount -o nobh' too along with it. Making the journal larger than
the mke2fs default often helps performance with metadata-intensive
workloads.




* writeback mode
In data=writeback mode, ext4 does not journal data at all. This mode provides a similar level of journaling as that of XFS, JFS, and ReiserFS in its default mode - metadata journaling. A crash+recovery can cause incorrect data to appear in files which were written shortly before the crash. This mode will typically provide the best ext4 performance.

[Yfrwlf]

http://cdimage.ubuntu.com/kubuntu/daily-live/current/

You could also wait for Alpha 3 which is arriving tomorrow, but tere shouldn't be any changes to this one really.

It's called system updates....doing that would probably be faster than downloading, burning/making a USB boot drive, and installing Alpha 3 when you already have Alpha 2.

Sillyfaces. :P

[broomfighter]

why haven't we been using xfs default?

[llama]

why haven't we been using xfs default?

It's tricky to use as a root FS, because you can only setup grub on it from the GRUB bootloader itself, not while running Linux with grub-install. (I keep a GRUB floppy image on a USB drive, that I can boot with syslinux's memdisk. I also have it on my PXE netboot server.)

Also, XFS needs to be mounted with logbsize=256k to do well all around (at least on magnetic disks). However, putting that in /etc/fstab for the root fs doesn't do anything, because for the root fs it only does a mount -o remount to try to apply the options. So if only logbsize=256k was the default on systems with plenty of RAM... (with the default logbsize=32k, IIRC, you bottleneck on journaling metadata ops any time you write a lot of small files, e.g. extracting a tar). Be sure you use a recent xfsprogs, or manually specify lazy_count=1.

Anyway, wow, ext4 does great on solid-state disks! I wonder what the bottlenecks are in the other filesystems... Maybe CPU usage on the Atom? I know XFS is not a small piece of code, and it's definitely tuned for massive SMP systems, but maybe not for systems with such a slow CPU, esp. relative to the I/O latency of a solid state disk.

[curaga]

JFS has the lowest CPU usage of all journaling FS..

[Yfrwlf]

I really like Ext4, very nice FS in comparison to Ext3, seems quite a lot faster. Conversion from Ext3 to 4 was easy and flawless. Rock on. ^^