Bug 1411215

Summary: CTDB NFS: With CTDB restart there are SELinux AVC's seen without any functional impact
Product: [Red Hat Storage] Red Hat Gluster Storage Reporter: surabhi <sbhaloth>
Component: gluster-nfsAssignee: Jiffin <jthottan>
Status: CLOSED WONTFIX QA Contact: Manisha Saini <msaini>
Severity: medium Docs Contact:
Priority: unspecified    
Version: rhgs-3.2CC: msaini, rcyriac, rhs-bugs, sanandpa, skoduri, storage-qa-internal
Target Milestone: ---Keywords: ZStream
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2018-11-19 04:14:41 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description surabhi 2017-01-09 06:42:45 UTC
Description of problem:
*************************************

When CTDB is restarted there are SELinux AVC's seen but there are no functional issues seen.
The failover also doesn't have any imapct.Raising a medium priority BZ for AVC's seen.


type=USER_AVC msg=audit(01/09/2017 12:06:11.143:3039632352) : pid=1 uid=root auid=unset ses=unset subj=system_u:system_r:init_t:s0 msg='avc:  denied  { stop } for auid=root uid=root gid=root path=/usr/lib/systemd/system/rpc-statd.service cmdline="/bin/systemctl stop nfslock.service" scontext=system_u:system_r:ctdbd_t:s0 tcontext=system_u:object_r:rpcd_unit_file_t:s0 tclass=service  exe=/usr/lib/systemd/systemd sauid=root hostname=? addr=? terminal=?' 
----
type=USER_AVC msg=audit(01/09/2017 12:06:11.143:3039632353) : pid=1 uid=root auid=unset ses=unset subj=system_u:system_r:init_t:s0 msg='avc:  denied  { status } for auid=root uid=root gid=root path=/usr/lib/systemd/system/rpc-statd.service cmdline="/bin/systemctl stop nfslock.service" scontext=system_u:system_r:ctdbd_t:s0 tcontext=system_u:object_r:rpcd_unit_file_t:s0 tclass=service  exe=/usr/lib/systemd/systemd sauid=root hostname=? addr=? terminal=?' 
----
type=USER_AVC msg=audit(01/09/2017 12:06:42.319:3039632356) : pid=1 uid=root auid=unset ses=unset subj=system_u:system_r:init_t:s0 msg='avc:  denied  { start } for auid=root uid=root gid=root path=/usr/lib/systemd/system/rpc-statd.service cmdline="/bin/systemctl start nfslock.service" scontext=system_u:system_r:ctdbd_t:s0 tcontext=system_u:object_r:rpcd_unit_file_t:s0 tclass=service  exe=/usr/lib/systemd/systemd sauid=root hostname=? addr=? terminal=?' 
----
type=SYSCALL msg=audit(01/09/2017 12:06:46.751:3039632360) : arch=x86_64 syscall=openat success=yes exit=3 a0=0xffffffffffffff9c a1=0x1af56f0 a2=O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC a3=0x0 items=0 ppid=1 pid=574 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=statd-callout exe=/usr/bin/bash subj=system_u:system_r:ctdbd_t:s0 key=(null) 
type=AVC msg=audit(01/09/2017 12:06:46.751:3039632360) : avc:  denied  { dac_override } for  pid=574 comm=statd-callout capability=dac_override  scontext=system_u:system_r:ctdbd_t:s0 tcontext=system_u:system_r:ctdbd_t:s0 tclass=capability 


Version-Release number of selected component (if applicable):
ctdb-4.4.6-2.el7rhgs.x86_64
glusterfs-3.8.4-10.el7rhgs.x86_64
selinux-policy-3.13.1-102.el7_3.11.noarch

How reproducible:
Always

Steps to Reproduce:
1.CTBD setup with nfs
2.Restart ctdb services
3.Observer audit logs

Actual results:
SELinux AVC's seen in audit logs .

Expected results:
There should not be any AVC's.

Additional info:

Since CTDB is not monitoring Kernel NFS (disabled) the statd-callout script trying to start stop and status for kernel nfs and it does not affect the functionality of RHGS CTDB nfs setup.

Comment 3 Niels de Vos 2017-01-09 09:48:34 UTC
From comment #0:

'/bin/systemctl stop nfslock.service' - this should not be done. Gluster/NFS does not use systemd to start the rpc.statd daemon. The process should probably get killed when Gluster/NFS is stopped though.

The openat() syscall should probably be allowed though, but only when Gluster/NFS has started rpc.statd and rpc.statd tries openat().

Comment 5 Niels de Vos 2017-01-09 09:53:41 UTC
This should be reported against the component that provides the ctdb shell script for monitoring/starting/stopping/... Gluster/NFS.