Am 07.10.2011 16:01, schrieb lst_ho...@kwsoft.de: >> Someone on the XFS mailinglist believed it could be filesystem >> fragmentation after all. They need an aligned continous 16k block to >> allocate a new inode chunk, otherwise it will fail. I'm going to test >> that later. > > This could be checked with "xfs_db -c frag -r /device" > > The second inode problem known with XFS is, that all inodes must be > below 1TB if you are not using inode64 mount option. But this should not > apply in your case :-)
Fragmentation looks good lxmhs45:~ # xfs_db -c frag -r /dev/sdb actual 1894079, ideal 1882306, fragmentation factor 0.62% However, the free extends histogramm requested by the ML does not. lxmhs45:~ # xfs_db -r "-c freesp -s" /dev/sdb from to extents blocks pct 1 1 32230 32230 5.36 2 3 6874 16476 2.74 4 7 138151 552604 91.90 total free extents 177255 total free blocks 601310 average free extent size 3.39234 So XFS is clearly having a problem here. I'm trying with ext4 now, let's see how it works out. Bernhard