Ceph

When executing mkcephfs on a new cluster the script throws the message "bash: line 0: cd: /home/matthew/forCeph: No such file or directory" (where "/home/matthew/forCeph" is whatever directory you run mkcephfs from). Creating the directory on each of the servers in the cluster eliminates the message, but no files are actually written to the directories. The cluster works fine after creation with or without the directories.

For example when executing mkcephfs from /home/matthew/forCeph part of the output is:

/usr/bin/monmaptool: writing epoch 0 to /tmp/mkcephfs.klxM1Uu9dO/monmap (3 monitors)
=== osd.0 === 
pushing conf and monmap to nut:/tmp/mkfs.ceph.19629
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
2012-03-22 19:43:46.821909 7f36dd108780 filestore(/srv/osd0) leveldb db created
2012-03-22 19:43:47.290829 7f36dd108780 created object store /srv/osd0 journal /srv/journal0/journal for osd.0 fsid 77ef45d8-c962-422d-9eca-ba4b3c546756
creating private key for osd.0 keyring /etc/ceph/osd.0.keyring
creating /etc/ceph/osd.0.keyring
collecting osd.0 key
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
=== osd.1 === 
pushing conf and monmap to geb:/tmp/mkfs.ceph.19629
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
2012-03-22 19:43:52.681163 7f9776c0e780 filestore(/srv/osd1) leveldb db created
2012-03-22 19:43:52.972825 7f9776c0e780 created object store /srv/osd1 journal /srv/journal1/journal for osd.1 fsid 77ef45d8-c962-422d-9eca-ba4b3c546756
creating private key for osd.1 keyring /etc/ceph/osd.1.keyring
creating /etc/ceph/osd.1.keyring
collecting osd.1 key
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
=== mds.a === 
pushing conf and monmap to nut:/tmp/mkfs.ceph.19629
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
bash: line 0: cd: /home/matthew/forCeph: No such file or directory
creating private key for mds.a keyring /etc/ceph/mds.a.keyring

After creating the directory "/home/matthew/forCeph" on each system in the cluster the errors are gone:

/usr/bin/monmaptool: writing epoch 0 to /tmp/mkcephfs.vHl68pFg11/monmap (3 monitors)
=== osd.0 === 
pushing conf and monmap to nut:/tmp/mkfs.ceph.20892
2012-03-22 19:45:13.933400 7fc783373780 filestore(/srv/osd0) leveldb db created
2012-03-22 19:45:14.402292 7fc783373780 created object store /srv/osd0 journal /srv/journal0/journal for osd.0 fsid 4001afc4-bf84-48ee-82d1-bfd70d30e77c
creating private key for osd.0 keyring /etc/ceph/osd.0.keyring
creating /etc/ceph/osd.0.keyring
collecting osd.0 key
=== osd.1 === 
pushing conf and monmap to geb:/tmp/mkfs.ceph.20892
2012-03-22 19:45:19.822337 7f297b71f780 filestore(/srv/osd1) leveldb db created
2012-03-22 19:45:20.113998 7f297b71f780 created object store /srv/osd1 journal /srv/journal1/journal for osd.1 fsid 4001afc4-bf84-48ee-82d1-bfd70d30e77c
creating private key for osd.1 keyring /etc/ceph/osd.1.keyring
creating /etc/ceph/osd.1.keyring
collecting osd.1 key
=== mds.a === 
pushing conf and monmap to nut:/tmp/mkfs.ceph.20892
creating private key for mds.a keyring /etc/ceph/mds.a.keyring