On Fri, 07 Jan 2005 18:54:21 +0100 Jaap Struyk <[EMAIL PROTECTED]> wrote:
> Op vr 07-01-2005, om 17:33 schreef Eric Jorgensen: > > > Where? I can't find it. Several pages on gentoo.org actually point > > to > > it's default support of xfs as one of the reasons it's better than > > redhat. Not that i have any interest in running gentoo. > > When you read the manual the "preparing the disks" section: > http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1&chap=4 > chapter "4.d. Creating Filesystems" > quote: > XFS is a filesystem with metadata journaling which comes with a robust > feature-set and is optimized for scalability. We only recommend using > this filesystem on Linux systems with high-end SCSI and/or fibre channel > storage and an uninterruptible power supply. Because XFS aggressively > caches in-transit data in RAM, improperly designed programs (those that > don't take proper precautions when writing files to disk and there are > quite a few of them) can lose a good deal of data if the system goes > down unexpectedly. Reiserfs is going to have the same problem, because it is also a metadata-only journaling system. If you pull the power, obviously, anything buffered disappears. SGI says: Q: Why do I see binary NULLS in some files after recovery when I unplugged the power? XFS journals metadata updates, not data updates. After a crash you are supposed to get a consistent filesystem which looks like the state sometime shortly before the crash, NOT what the in memory image looked like the instant before the crash. Since XFS does not write data out immediately unless you tell it to with fsync or an O_SYNC open (the same is true of other filesystems), you are looking at an inode which was flushed out, but whose data was not. You will find that the inode is not taking any space since all it has is a size but no extents allocated (run xfs_bmap(8) on the file). But clearly this should only be in files which were open for writing at the time of failure. So, were you in the middle of an ebuild or what? On the other hand, a drive that is writing during an unclamped spike in power or during a brownout may well hit the wrong physical sector on the disk. The theory (and sometimes practice) is that server-oriented drives, like scsi and FC drives, should be designed to avoid doing dangerous things in the event of a power loss, whether it's clean or dirty, where consumer-oriented drives probably aren't even aware that power is failing until it's time to park the heads. Having a computer boot back up immediately as power is restored is also a very good way to kill it, as mains power will fluctuate wildly for the first several seconds after it comes back up while the grid reacts to the load. Running a computer you care about without a UPS is a questionable practice anyhow. They aren't that expensive, compared to replacing the computer, or the time required to rebuild it. I don't buy it. I don't think xfs is intrinsically less stable than any other filesystem. Four years ago people were asking me why i wasn't worried about filesystem corruption when i told them i was using reiserfs, and somehow i never experienced any. But i have a personal bias against the gentoo team, their fans, and that ilk in general, so it has less credibility coming from them. ------------------------------------------------------- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt _______________________________________________ Freevo-users mailing list Freevo-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freevo-users
