Red Hat Bugzilla – Bug 1158571
user and group support in gui - permissions to clusters managed by pcsd
Last modified: 2015-11-19 04:33:19 EST
Add ability to set read only/write/admin access to each cluster managed by pcsd instance.
Created attachment 1061040 [details] proposed fix
Before Fix: [root@rh71-node1 ~]# rpm -q pcs pcs-0.9.142-2.el7.x86_64 No support for users and groups permissions to clusters in web UI apart from pacemakers ACLs. After Fix: [root@rh71-node1:~]# rpm -q pcs pcs-0.9.143-1.el7.x86_64 There is a new tab "Permissions" on the main page of the web UI. It allows to view and edit permissions for clusters managed by the web UI. It is possible to grant permissions for each cluster to users and groups, meaning of each permission is explained in a tooltip. Permissions are enforced throughout the web UI. If user tries to execute an action which is not allowed for him/her to run, an error message is displayed.
Created attachment 1070256 [details] proposed fix 2
Created attachment 1070281 [details] proposed fix 3
Before Fix: [root@rh71-node1 ~]# rpm -q pcs pcs-0.9.143-6.el7.x86_64 Web UI does not inform about unsaved changes in a cluster permissions form. Pcsd checks user's group membership upon login only. It does not react on changes done while user is logged in. After Fix: [root@rh71-node1:~]# rpm -q pcs pcs-0.9.143-7.el7.x86_64 Web UI informs about unsaved changes in a cluster permissions form by displaying an asterisk next to the cluster name. Let's have user user1 who is member of adm and haclient groups. Set permissions in pcsd to: user user1 - read group adm - full User1 is able to log in to web UI and change cluster settings. If user1's membership in adm group is revoked while the user is logged in, the user is no longer able to change the cluster settings. If user1's membership in haclient group is revoked, a login dialog is displayed to the user.
Created attachment 1071049 [details] proposed fix 4
Created attachment 1071234 [details] proposed fix 5 Let's have a cluster consisting from rh71-node1 and rh71-node2, and a web UI host rh71-node3. [root@rh71-node1:~]# id user1 uid=1000(user1) gid=1000(user1) groups=1000(user1) [root@rh71-node2:~]# id user1 uid=1000(user1) gid=1000(user1) groups=1000(user1) [root@rh71-node3:~]# id user1 uid=1000(user1) gid=1000(user1) groups=1000(user1),189(haclient) Before Fix: [root@rh71-node1 ~]# rpm -q pcs pcs-0.9.143-7.el7.x86_64 Status of the cluster in the web UI is "unknown" and it id not possible do view nor change its settings. After Fix: [root@rh71-node1:~]# rpm -q pcs pcs-0.9.143-8.el7.x86_64 Log in to web UI on node3 as user1. You are able to manage the cluster.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHSA-2015-2290.html