Assuming neither the pool nor volume NDS objects have been created i'd start checking the "NSS identity". By default an user object named "servernameadmin" (if the servername is "FS1" the object's name would be "FS1admin") with supervisory rights the the server's context gets created when NSS gets initially configured. Further on this object serves as identity for stuff such as creating volumes. There's no referential integrity for this object, so if it gets deleted / renamed / moved (or it doesn't have the needed rights or the server object gets moved...) the behaviour you encounter would be expected. In such cases you could simply rerun the NSS config workflow in yast.