heavy NFS writes lead to corrup summary in superblock

[Mikhail Teterin]

п'ятниця 09 червень 2006 02:56, R. B. Riddick написав:
> I say, does that discrepancy persist, when you just wait some time?

Yes... I'm noticing this hours after the dumps ended...

> I would guess, that something has an open file descriptor on a deleted
> file, so that this file cannot be really deleted (it just disappears from
> the directory tree)...

If anything did, I wouldn't be able to umount the filesystem cleanly, would I? 
Yet, it unmounts peacefully, even though the subsequent fsck finds the 
superblock summary to be incorrect.

When I tried to use the FS as a scratch for an unrelated thing, though, I 
noticed some processes hanging in nbufkv state. Google-ing led me to the:

	http://lists.freebsd.org/pipermail/freebsd-current/2003-June/004702.html

Is this 3 year old advise *still* true? I rebuilt the kernel with BKVASIZE 
bumped to 64K (the block size on the FS in question) and am running another 
batch of dumps right now. When it is over, I'll check the df/du...

Yours,

	-mi