I asked the XFS folks who mentioned that the issues with 64 bit inodes are old, constrained to larger filesystems than what I'm using, not an issue with nfsv4, and not present on anything but 32bit clients with old userspace. In any case, I have been experimenting a bit and somehow the issue seems to be related to exporting with sec=krb5i:krb5p or sec=krb5i. If I export with just sec=krb5p, things magically begin to work. So basically: [root@ld00 ~]# ls -l ~tester|wc -l; grep tester /proc/mounts 7685 nas00:/export/misc-00/tester /home/tester nfs4 rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=krb5p,clientaddr=172.21.84.191,local_lock=none,addr=172.21.86.77 0 0 (unmount, then re-export with krb5i on the server) [root@ld00 ~]# ls -l ~tester|wc -l; grep tester /proc/mounts ls: reading directory '/home/tester': Input/output error 5623 nas00:/export/misc-00/tester /home/tester nfs4 rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=krb5i,clientaddr=172.21.84.191,local_lock=none,addr=172.21.86.77 0 0 (umount, then re-export with krb5i:krb5p on the server) [root@ld00 ~]# ls -l ~tester|wc -l; grep tester /proc/mounts ls: reading directory '/home/tester': Input/output error 5623 nas00:/export/misc-00/tester /home/tester nfs4 rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=krb5i,clientaddr=172.21.84.191,local_lock=none,addr=172.21.86.77 0 0 (umount, switch back to plain krb5p) [root@ld00 ~]# ls -l ~tester|wc -l; grep tester /proc/mounts 7685 nas00:/export/misc-00/tester /home/tester nfs4 rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=krb5p,clientaddr=172.21.84.191,local_lock=none,addr=172.21.86.77 0 0 Sometimes the number of files it lists before it fails changes (and in this case has been as small as a few hundred) but I don't know what causes it to change. Anyway, I hope this helps to pinpoint the problem. I now have a really easy way to reproduce this without having to kick people off of the server, and if the successes aren't just some kind of false positives then I guess I also have a workaround. I'm still at a loss as to why a revert of the readdir changes makes any difference at all here. - J<
