There are different ways. It all depends on the authentication is set and user mapping rules. For example, in UNIX, security is set based on UID/GID. In windows, the SID is mapped to UID and GID respectively using a Usermapper feature. If the cluster has two authentication sources and the user logs in via AD, the POSIX mapper will attempt a match with the user id and get it's respective UID. If the userid does not exsit, then it will pull a UID from a pool that is stored locally and starts at 1MM. If you have authentication issues, clearing the user mapper database will remove all mappings.
PaulDanis
13 Posts
439
0
Posted September 10th, 2013 08:00
There are different ways. It all depends on the authentication is set and user mapping rules. For example, in UNIX, security is set based on UID/GID. In windows, the SID is mapped to UID and GID respectively using a Usermapper feature. If the cluster has two authentication sources and the user logs in via AD, the POSIX mapper will attempt a match with the user id and get it's respective UID. If the userid does not exsit, then it will pull a UID from a pool that is stored locally and starts at 1MM. If you have authentication issues, clearing the user mapper database will remove all mappings.