Using the new mkfs.xfs, mongo runs smooth now. Thanks! :)
Steve Lord wrote:
Just in case I did not make it obvious, we think the fix for this is a
new mkfs. The old one was generating inconsistent superblocks for
filesystems above 1 Tbyte. The only folks affected by this probably
had filesystems which would take down the machine if you used them
anyway. So if your filesystem is working for you using the old mkfs
then there is no need to do anything except upgrade the commands.
If you made a filesystem of greater than 1 Tbyte using
mkfs -t xfs -i size=512 .....
then you should also not be affected.