Skip to contentSkip to conversation

XFS?

DISCUSSION Martin Knoblauch 4 participants 5 messages Sep 12, 2002 – Sep 13, 2002

lore
Share & tools
lore
#0 lore raw
Quoted 8 lines
Quoted 4 lines

So does Redhat/Suse/??? ship XFS yet?

john

Mandrake has had XFS support in the default boot kernel since 8.0. AFAIK, Suse
and Slackware also have XFS capable kernels now too.

for what its worth, MSC.Linux supports it on IA32 and IA64 :-)

In my opinion the non-inclosure in the mainline kernel is the most
important reason not to use XFS (or any other FS). Which in turn massively reduces the tester base. It is a shame, because for some type of applications it performs great, or better than anything else.

Martin

Signature · 10 lines
Martin Knoblauch
Senior System Architect
MSC.software GmbH
Am Moosfeld 13
D-81829 Muenchen, Germany

e-mail: [email protected]
http://www.mscsoftware.com
Phone/Fax: +49-89-431987-189 / -7189
Mobile: +49-174-3069245
#1 lore raw
Quoted 4 lines

In my opinion the non-inclosure in the mainline kernel is the most important reason not to use XFS (or any other FS). Which in turn massively reduces the tester base. It is a shame, because for some type of applications it performs great, or better than anything else.

On the other hand, filesystem corruption bugs are one of the worst type to suffer from. We absolutely don't want to include filesystems without at least a reasonable proven track record in the mainline kernel, and therefore encourage the various distributions to use them, incase any bugs do show up. Look how long a buffer overflow existed in Zlib unnoticed.

EXT2 is a very capable filesystem, and has *years* of proven reliability. That's why I'm not going to switch away from it for critical work any time soon.

John.

#2 lore raw
Quoted 6 lines · On Thu, 12 Sep 2002 [email protected] wrote:
Quoted 4 lines

In my opinion the non-inclosure in the mainline kernel is the most important reason not to use XFS (or any other FS). Which in turn massively reduces the tester base. It is a shame, because for some type of applications it performs great, or better than anything else.

Quoted 6 lines

On the other hand, filesystem corruption bugs are one of the worst type to suffer from. We absolutely don't want to include filesystems without at least a reasonable proven track record in the mainline kernel, and therefore encourage the various distributions to use them, incase any bugs do show up. Look how long a buffer overflow existed in Zlib unnoticed.

Given that the IDE code in 2.5 wrote random bad data not only in the mounted filesystems but on other partitions and even drives, if we are dropping things which have an unreasonable track record, we should drop IDE for sure ;-)

This is a development kernel, the rules for what goes in should be far more open than the stable series. IMHO both JFS (AIX) and XFS (IRIX) should be in, because they will not be solid until users actually use them, and better that be in a development kernel.

Quoted 4 lines

EXT2 is a very capable filesystem, and has *years* of proven reliability. That's why I'm not going to switch away from it for critical work any time soon.

One might note that both JFS and XFS have been around since xiafs was the Linux f/s of choice. It's all relative. If you want old and grotty, go back to minix f/s.

Signature · 3 lines
bill davidsen <[email protected]>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
#3 lore raw
Quoted 13 lines
Quoted 6 lines
Quoted 4 lines

In my opinion the non-inclosure in the mainline kernel is the most important reason not to use XFS (or any other FS). Which in turn massively reduces the tester base. It is a shame, because for some type of applications it performs great, or better than anything else.

Quoted 6 lines

On the other hand, filesystem corruption bugs are one of the worst type to suffer from. We absolutely don't want to include filesystems without at least a reasonable proven track record in the mainline kernel, and therefore encourage the various distributions to use them, incase any bugs do show up. Look how long a buffer overflow existed in Zlib unnoticed.

Quoted 4 lines

Given that the IDE code in 2.5 wrote random bad data not only in the mounted filesystems but on other partitions and even drives, if we are dropping things which have an unreasonable track record, we should drop IDE for sure ;-)

Things have certainly changed, (for better or worse, I'm not sure), since the 1.3.X days when a development kernel was generally still pretty stable.

Quoted 4 lines

This is a development kernel, the rules for what goes in should be far more open than the stable series. IMHO both JFS (AIX) and XFS (IRIX) should be in, because they will not be solid until users actually use them, and better that be in a development kernel.

Totally agreed. I was talking about the stable kernel.

Quoted 6 lines
Quoted 3 lines

EXT2 is a very capable filesystem, and has *years* of proven reliability. That's why I'm not going to switch away from it for critical work any time soon.

One might note that both JFS and XFS have been around since xiafs was the Linux f/s of choice.

Not for Linux, though - I'm talking about years of Linux stability.

Quoted 1 line

It's all relative. If you want old and grotty, go back to minix f/s.

That's why I qualified my above comment with 'is a very capable filesystem' :-).

I know what you mean, but I was just pointing out that EXT-2 balances proven reliability in the stable kernel, features, and performance VERY well, infact what other OS family can make that claim? BSD is the only one I can think of. Oh, sure FAT has been around forever, but it's somewhat lacking in the features department.

John.

#4 lore raw
In reply to Bill Davidsen
Quoted 4 lines · Bill Davidsen

This is a development kernel, the rules for what goes in should be far more open than the stable series. IMHO both JFS (AIX) and XFS (IRIX) should be in, because they will not be solid until users actually use them, and better that be in a development kernel.

JFS is in isnt it?

Quoted 3 lines

One might note that both JFS and XFS have been around since xiafs was the Linux f/s of choice. It's all relative. If you want old and grotty, go back to minix f/s.

Ah, you remember xiafs. Perhaps the first big flamewar on the kernel mailing list.

Jim