On 1 May 2002, Eric Sandeen wrote: > Just for kicks you could try using xfs_growfs to bump the inode maxpct a > bit, although from df -i I don't think that's the problem... Do you happen to know if maxpct inodes is based upon size of the fs or just a constant? 12 could be interpreted as 12.5% which would be exactly half. And that sounds fishy to me. Cheers Seth