Bug 1194000

Summary: Nothing starts rpcbind.socket
Product: [Fedora] Fedora Reporter: Jason Tibbitts <j>
Component: systemdAssignee: systemd-maint
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 21CC: bfields, jlayton, johannbg, jsynacek, lnykryn, msekleta, s, steved, systemd-maint, teg, t.h.amundsen, vpavlin, zbyszek
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-07-01 19:40:16 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 Jason Tibbitts 2015-02-18 18:05:16 UTC
I've been trying to debug a problem where on a fresh F21 install, I can't do an NFS3 mount.  Mounting tries to start rpc.statd (rpc-statd.service) which requires rpcbind.target, and rpcbind.target actually starts fine.  But rpcbind is socket activated and nothing at all that I can see actually starts rpcbind.socket.

If I start rpcbind.socket manually, everything works fine.

rpcbind-0.2.2-2.1.fc21.x86_64
nfs-utils-1.3.1-6.1.fc21.x86_64
systemd-216-20.fc21.x86_64

Comment 1 Jason Tibbitts 2015-02-18 18:56:59 UTC
And, in case anyone else stumbles on this:

systemctl enable rpcbind.socket
systemctl start rpcbind.socket

works around the problem well enough, though from what I gather this isn't supposed to be necessary.

Comment 2 Steve Dickson 2015-02-26 22:43:47 UTC
I'm beginning to thinks this is a systemd issued. Ever since

commit 061eaaeeb588122350fa3fc4d81ac814c59fbd84
Author: Tom Gundersen <teg>
Date:   Tue Nov 25 11:41:50 2014 -0500

    rpcbind: add support for systemd socket activation

There has been problems. 

I noticed today that if you reboot a machine and the
first thing done is a 'rpcinfo -p'. The command hangs
but rpcbind is started! The rpcinfo will time out 
complaining about not being able to talk to rpcbind.

If another rpcinfo -p is done, it works as expected. 

So it appears to be a problem in the systemd activation 
code or the implementation of the systemd activation 
code.

I'm assigning this to the systemd folks because I simply
have no idea how to debug this

Comment 3 Jason Tibbitts 2015-06-30 22:13:27 UTC
As far as I can tell, this problem has gone away.

I removed the bits from my kickstart files which force-enabled rpcbind.socket and installed fresh F21 and F22 machines.  Everything seems to be working as expected.

Comment 4 Steve Dickson 2015-07-01 19:40:16 UTC
(In reply to Jason Tibbitts from comment #3)
> As far as I can tell, this problem has gone away.
> 
> I removed the bits from my kickstart files which force-enabled
> rpcbind.socket and installed fresh F21 and F22 machines.  Everything seems
> to be working as expected.

A patch went in for nfs-utils-1.3.2-9.fc22 and nfs-utils-1.3.1-6.4.fc21
that took care of this problem.

*** This bug has been marked as a duplicate of bug 1178720 ***