[Samba] Howto compute (calculate) RID while using the LDAP backend
Karel Lang AFD
lang at afd.cz
Mon Aug 25 11:03:23 MDT 2014
Hello,
as my subject says, i have a few questions regarding how to calculate
the RID part of SID, or how to value should look like..
few facts 1st:
- OS is RHEL 6.5
- samba version:
samba.x86_64 3.6.9-164.el6
samba-client.x86_64 3.6.9-164.el6
samba-common.x86_64 3.6.9-164.el6
samba-winbind.x86_64 3.6.9-164.el6
samba-winbind-clients.x86_64 3.6.9-164.el6
samba4-libs.x86_64 4.0.0-58.el6.rc4
- smbldap tools:
smbldap-tools.noarch 0.9.10-1.el6
- ldap backend (389 directory server)
389-ds-base.x86_64 1.2.11.15-29.el6
My question is:
1. why there is difference in RID value while adding user to LDAP
backend through the smbldap script "smbldap-useradd" versus "smbpasswd -a" ?
2. i'd like to have posix UID equal to RID - is this feasible?
my UID all start above 1000
Please now let me elaborate to more detail to fig 1:
"smbldap-useradd" produce RID based on equation: UID*2 + 1000 resulting
my user having UID 1001 will have RID 3002 stored in LDAP backend
"smbpasswd -a" produce next free RID starting also at 1000 - so eg. if i
add user via 389-console to LDAP and specify the user account posix
attributes and the i use "smbpasswd -a" to add Samba attributes, the
resulting RID for posix user with UID 1001 would be again 1001 (or the
next free)
This discrepancy bothers me a bit, because i'd like to have possibility
to add users both way to LDAP while keeping the RID numbering tidy and
logical.
With this being said, i'd prefer (for human readability reasons) to keep
RID equal to UID - as i asked in my fig. 2 but basically i'm not sure
why the smbldap-tools devs came up with that calculation equation?
For sure for some reason?
Anyone been in same boat wandering about what RID values should be?
Could anyone share some light onto this matter for me?
Big thanks!
Karel Lang
lang 'at' afd.cz
More information about the samba
mailing list