> Looks like there has been some 'discussion' over the weekend. There have
I'd put it without the quotes... it was useful to me, anyway.
> o There was a stall problem where the raid would just grind to a halt,
> this was fixed by a kernel change in the 2.4.7-pre series. This was
> in the 1.0.1 release and would exhibit itself as a filesystem which
> just ground to a halt.
That's the one I couldn't find in the list; thanks.
> o The second was a problem writing out an internal log. The log was
> not getting written correctly on raid5 devices, which meant you
> could not always mount the filesystem again without running
> xfs_repair. This problem was introduced and fixed in the cvs tree
> after the 1.0.1 release.
After the weekend's discussion, I had a `clever' idea. Each disk
in the array has a 128 MB partition, and a ~180 GB partition. The former
are in a raid1 array with the log, the latter raid5 with the main XFS.
Assuming this is a correct solution, it answers all my concerns;
Everything is on the same array, and moving it means moving the raidtab
and fstab entries, which one would have to do, anyway. And, of course,
it's at least as safe as raid5 by itself.
> As for raid5 log performance, the log is written in chunks of upto 32Kbytes,
> but these chunks can be any 512 byte multiple long, and start on any 512
> byte boundary. This does not appear to be well treated by the raid5 code
> and results in less than optimal performance. A raid1 external log works
For our use pattern (avg file size ~1GB, near-WORM usage) performance
is not nearly the issue that stability and safety are. I'll care a lot more
about this when we get round to using it for home directories.
> It may be that your problem is still there and is being caused by the
> layout of the xfs logwrites, in which case the only fix may be to move
> to an external log until we can do something about the alignment of the
> log writes.
Well, we have one system loaded up with a TB of data, which can
just be mounted read-only until we're done processing it, and the other
system now has an external log, since it wasn't in production yet.
Also, it doesn't look to me like XFS is doing anything illegal;
linux raid5 is kinda new as far as the 2.4 kernels are concerned...
Just to be sure... thanks, Seth and Steve, for the discussion 8)
----
Chris J. Bednar <http://optics.tamu.edu/~bednar/>
Director, Distributed Computing Product Group
http://AdvancedDataSolutions.com/
|