Bug 14254 - Have to reconfig and restart the nis server after every reboot
Summary: Have to reconfig and restart the nis server after every reboot
Status: NEW
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
Priority: High major
Target Milestone: ---
Assignee: Joshua Santos
QA Contact:
URL:
Whiteboard: MGA5TOO
Keywords: PATCH, Triaged
Depends on:
Blocks:
 
Reported: 2014-10-09 05:47 CEST by Robert Courtright
Modified: 2016-10-10 18:08 CEST (History)
7 users (show)

See Also:
Source RPM: ypserv-2.32.1-4.mga5
CVE:
Status comment:


Attachments

Description Robert Courtright 2014-10-09 05:47:23 CEST
Description of problem:
Have to reconfig and restart the nis server after every reboot

Version-Release number of selected component (if applicable):
Cauldron

How reproducible:
Every reboot

Steps to Reproduce:
1. Install ypserv and config
2. Check client for service.
3. reboot server and check clients


Reproducible: 

Steps to Reproduce:
Robert Courtright 2014-11-30 18:33:54 CET

Target Milestone: --- => Mageia 5

Comment 1 Robert Courtright 2014-12-20 23:02:17 CET
is anyone looking into this bug?
Comment 2 Robert Courtright 2015-01-26 16:37:26 CET
Well I no longer have to re-config. Just start it every time the system is booted. I have check it is set to start at boot.
Comment 3 Stephane 2015-01-27 22:52:18 CET
I installed ypserv without configuration just to test if the service start with boot on Mageia5 beta2.

Just after the installation the service is not started which seems normal because there is no need to run it without first configuration.
But after a reboot the service is running.

Can you do a new test with Mageia5 beta2 ?

These output after the reboot :

# systemctl status ypserv.service 
â ypserv.service - LSB: The ypserver daemon (for a NIS/YP server)
   Loaded: loaded (/etc/rc.d/init.d/ypserv)
   Active: active (exited) since mar. 2015-01-27 22:06:39 CET; 1min 0s ago
  Process: 932 ExecStart=/etc/rc.d/init.d/ypserv start (code=exited, status=0/SUCCESS)

janv. 27 22:06:39 localhost ypserv[932]: Lancement du serveur YP : [  OK  ]

CC: (none) => stephane_oss

Comment 4 Robert Courtright 2015-01-28 17:35:49 CET
This is what I get when I do the same:

# systemctl status ypserv.service
â ypserv.service - LSB: The ypserver daemon (for a NIS/YP server)
   Loaded: loaded (/etc/rc.d/init.d/ypserv)
   Active: inactive (dead)


This is beta 2 with all updates.
Once I start it I get:
# systemctl status ypserv.service
â ypserv.service - LSB: The ypserver daemon (for a NIS/YP server)
   Loaded: loaded (/etc/rc.d/init.d/ypserv)
   Active: active (running) since Wed 2015-01-28 10:33:11 CST; 2s ago
  Process: 10205 ExecStart=/etc/rc.d/init.d/ypserv start (code=exited, status=0/SUCCESS)
   CGroup: /system.slice/ypserv.service
           ââ10215 ypserv

Jan 28 10:33:11 hostname systemd[1]: Starting LSB: The ypserver daemon (for a NIS/YP server)...
Jan 28 10:33:11 hostname ypserv[10215]: WARNING: no securenets file found!
Jan 28 10:33:11 hostname ypserv[10205]: Starting YP server services: [  OK  ]
Jan 28 10:33:11 hostname systemd[1]: Started LSB: The ypserver daemon (for a NIS/YP server).

Of course hostname replace by system name.
I can now log onto client machines
Comment 5 Robert Courtright 2015-01-28 17:36:33 CET
Oh by the way I have to set the domain name with every boot.
Comment 6 Robert Courtright 2015-02-06 14:21:03 CET
here some more information as to why it won start. From the boot log.

SKIP ^[[0m] Ordering cycle found, skipping LSB:...ver daemon (for a NIS/YP server)
SKIP ^[[0m] Ordering cycle found, skipping LSB:...frd daemon (for a NIS/YP server)
Comment 7 Samuel Verschelde 2015-05-20 10:29:19 CEST
Unfortunately this package has no dedicated maintainer. Adding 2 packagers who updated the package at some point in the past, and also Colin who might want to give some insight about the systemd issue.

CC: (none) => luigiwalser, mageia, thierry.vignaud

Samuel Verschelde 2015-05-20 10:29:25 CEST

Keywords: (none) => Triaged

Samuel Verschelde 2015-05-20 10:29:42 CEST

Whiteboard: (none) => MGA5TOO FOR_ERRATA

Samuel Verschelde 2015-06-08 17:14:05 CEST

Target Milestone: Mageia 5 => ---
Priority: Normal => High
Whiteboard: MGA5TOO FOR_ERRATA => MGA5TOO

Comment 8 Joshua Santos 2015-08-12 03:41:37 CEST
Hi,
  My name is Josh, and I am new to the packaging team, and do not have a mentor yet, however I would like to take this project on :)

Regards

CC: (none) => nerdsville

Joshua Santos 2015-08-12 03:42:57 CEST

Assignee: bugsquad => nerdsville

Comment 9 David Walser 2015-08-12 03:43:17 CEST
Hi Josh,

Glad to see you here.  If you find a solution for this bug, please post it and we'll have a look.
Comment 10 Joshua Santos 2015-08-12 05:41:45 CEST
So I have configured ypserv on mga 5, and it appears the settings are persisting:

[root@localhost nerdsville]# service ypserv status
â ypserv.service - LSB: The ypserver daemon (for a NIS/YP server)
   Loaded: loaded (/etc/rc.d/init.d/ypserv)
   Active: active (exited) since Tue 2015-08-11 23:37:28 EDT; 54s ago
  Process: 954 ExecStart=/etc/rc.d/init.d/ypserv start (code=exited, status=0/SUCCESS)

Aug 11 23:37:28 localhost.localdomain ypserv[954]: Starting YP server services: [  OK  ]

One thing I do notice, however, is that the default settings are never copied into their appropriate directories, could this be the issue we are looking at? The settings as I configured them seem to persist through reboot as it stands, but I did need to copy over the files manually, the package installation did not do this.

Regards
Comment 11 Joshua Santos 2015-08-12 06:02:37 CEST
Seems to start at boot as well, maybe this has already been fixed?
Comment 12 Joshua Santos 2015-08-12 18:06:03 CEST
(In reply to Robert  Courtright from comment #6)
> here some more information as to why it won start. From the boot log.
> 
> SKIP ^[[0m] Ordering cycle found, skipping LSB:...ver daemon (for a NIS/YP
> server)
> SKIP ^[[0m] Ordering cycle found, skipping LSB:...frd daemon (for a NIS/YP
> server)

Do you think you could please provide a list of the services that are running under systemd so that I could mock the environment?

This command should provide this information: systemctl list-units
Comment 13 andre salaun 2015-09-24 00:48:56 CEST
(In reply to Joshua Santos from comment #12)
> (In reply to Robert  Courtright from comment #6)
> > here some more information as to why it won start. From the boot log.
> > 
> > SKIP ^[[0m] Ordering cycle found, skipping LSB:...ver daemon (for a NIS/YP
> > server)
> > SKIP ^[[0m] Ordering cycle found, skipping LSB:...frd daemon (for a NIS/YP
> > server)
> 
> Do you think you could please provide a list of the services that are
> running under systemd so that I could mock the environment?
> 
> This command should provide this information: systemctl list-units

I confirm (at home, i just installed ypserv) lost domainname on boot (writed in /etc/sysconfig/network)
I have in boot.log :

[ SKIP ] Ordering cycle found, skipping LSB: The ypserver daemon (for a NIS/YP server)

and failure on LSB :

systemctl status network.service
â network.service - LSB: Bring up/down networking
   Loaded: loaded (/etc/rc.d/init.d/network)
   Active: failed (Result: exit-code) since jeu. 2015-09-24 00:24:48 CEST; 10min ago
  Process: 1683 ExecStart=/etc/rc.d/init.d/network start (code=exited, status=1/FAILURE)
   CGroup: /system.slice/network.service
           ââ1860 /sbin/ifplugd -I -b -i eno1

sept. 24 00:24:48 gaia.terre systemd[1]: network.service: control process exited, code=exited status=1
sept. 24 00:24:48 gaia.terre systemd[1]: Failed to start LSB: Bring up/down networking.
sept. 24 00:24:48 gaia.terre systemd[1]: Unit network.service entered failed state.
sept. 24 00:24:48 gaia.terre systemd[1]: network.service failed.
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using interface eno1/74:D4:35:BD:31:BB with driver <e1000e> (version: 2.3.2-k)
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using detection mode: SIOCETHTOOL
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Initialization complete, link beat not detected.
sesystemctl status network.service
â network.service - LSB: Bring up/down networking
   Loaded: loaded (/etc/rc.d/init.d/network)
   Active: failed (Result: exit-code) since jeu. 2015-09-24 00:24:48 CEST; 10min ago
  Process: 1683 ExecStart=/etc/rc.d/init.d/network start (code=exited, status=1/FAILURE)
   CGroup: /system.slice/network.service
           ââ1860 /sbin/ifplugd -I -b -i eno1

sept. 24 00:24:48 gaia.terre systemd[1]: network.service: control process exited, code=exited status=1
sept. 24 00:24:48 gaia.terre systemd[1]: Failed to start LSB: Bring up/down networking.
sept. 24 00:24:48 gaia.terre systemd[1]: Unit network.service entered failed state.
sept. 24 00:24:48 gaia.terre systemd[1]: network.service failed.
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using interface eno1/74:D4:35:BD:31:BB with driver <e1000e> (version: 2.3.2-k)
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using detection mode: SIOCETHTOOL
sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Initialization complete, link beat not detected.
sept. 24 00:24:52 gaia.terre ifplugd(eno1)[1860]: Link beat detected.
sept. 24 00:24:53 gaia.terre ifplugd(eno1)[1860]: Executing '/etc/ifplugd/ifplugd.action eno1 up'.
sept. 24 00:24:56 gaia.terre ifplugd(eno1)[1860]: Program executed successfully.
pt. 24 00:24:52 gaia.terre ifplugd(eno1)[1860]: Link beat detected.
sept. 24 00:24:53 gaia.terre ifplugd(eno1)[1860]: Executing '/etc/ifplugd/ifplugd.action eno1 up'.
sept. 24 00:24:56 gaia.terre ifplugd(eno1)[1860]: Program executed successfully.

On another ypserver in my network school (upgrade from mageia 4), ypserv keep is domainname after reboot but client users can not connect with ypbind (some clients run mageai4 others mageia 5). So I think something have changed or something is wrong in ypserv.

Looking for a bug I notice there is an error on clients with setsebool (does not find file) of selinux when starting ypbind. However selinux is not installed nor set.

CC: (none) => andresalaun

Comment 14 andre salaun 2015-09-24 19:06:46 CEST
(In reply to andre salaun from comment #13)
> (In reply to Joshua Santos from comment #12)
> > (In reply to Robert  Courtright from comment #6)
> > > here some more information as to why it won start. From the boot log.
> > > 
> > > SKIP ^[[0m] Ordering cycle found, skipping LSB:...ver daemon (for a NIS/YP
> > > server)
> > > SKIP ^[[0m] Ordering cycle found, skipping LSB:...frd daemon (for a NIS/YP
> > > server)
> > 
> > Do you think you could please provide a list of the services that are
> > running under systemd so that I could mock the environment?
> > 
> > This command should provide this information: systemctl list-units
> 
> I confirm (at home, i just installed ypserv) lost domainname on boot (writed
> in /etc/sysconfig/network)
> I have in boot.log :
> 
> [ SKIP ] Ordering cycle found, skipping LSB: The ypserver daemon (for a
> NIS/YP server)
> 
> and failure on LSB :
> 
> systemctl status network.service
> â network.service - LSB: Bring up/down networking
>    Loaded: loaded (/etc/rc.d/init.d/network)
>    Active: failed (Result: exit-code) since jeu. 2015-09-24 00:24:48 CEST;
> 10min ago
>   Process: 1683 ExecStart=/etc/rc.d/init.d/network start (code=exited,
> status=1/FAILURE)
>    CGroup: /system.slice/network.service
>            ââ1860 /sbin/ifplugd -I -b -i eno1
> 
> sept. 24 00:24:48 gaia.terre systemd[1]: network.service: control process
> exited, code=exited status=1
> sept. 24 00:24:48 gaia.terre systemd[1]: Failed to start LSB: Bring up/down
> networking.
> sept. 24 00:24:48 gaia.terre systemd[1]: Unit network.service entered failed
> state.
> sept. 24 00:24:48 gaia.terre systemd[1]: network.service failed.
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using interface
> eno1/74:D4:35:BD:31:BB with driver <e1000e> (version: 2.3.2-k)
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using detection mode:
> SIOCETHTOOL
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Initialization complete,
> link beat not detected.
> sesystemctl status network.service
> â network.service - LSB: Bring up/down networking
>    Loaded: loaded (/etc/rc.d/init.d/network)
>    Active: failed (Result: exit-code) since jeu. 2015-09-24 00:24:48 CEST;
> 10min ago
>   Process: 1683 ExecStart=/etc/rc.d/init.d/network start (code=exited,
> status=1/FAILURE)
>    CGroup: /system.slice/network.service
>            ââ1860 /sbin/ifplugd -I -b -i eno1
> 
> sept. 24 00:24:48 gaia.terre systemd[1]: network.service: control process
> exited, code=exited status=1
> sept. 24 00:24:48 gaia.terre systemd[1]: Failed to start LSB: Bring up/down
> networking.
> sept. 24 00:24:48 gaia.terre systemd[1]: Unit network.service entered failed
> state.
> sept. 24 00:24:48 gaia.terre systemd[1]: network.service failed.
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using interface
> eno1/74:D4:35:BD:31:BB with driver <e1000e> (version: 2.3.2-k)
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Using detection mode:
> SIOCETHTOOL
> sept. 24 00:24:48 gaia.terre ifplugd(eno1)[1860]: Initialization complete,
> link beat not detected.
> sept. 24 00:24:52 gaia.terre ifplugd(eno1)[1860]: Link beat detected.
> sept. 24 00:24:53 gaia.terre ifplugd(eno1)[1860]: Executing
> '/etc/ifplugd/ifplugd.action eno1 up'.
> sept. 24 00:24:56 gaia.terre ifplugd(eno1)[1860]: Program executed
> successfully.
> pt. 24 00:24:52 gaia.terre ifplugd(eno1)[1860]: Link beat detected.
> sept. 24 00:24:53 gaia.terre ifplugd(eno1)[1860]: Executing
> '/etc/ifplugd/ifplugd.action eno1 up'.
> sept. 24 00:24:56 gaia.terre ifplugd(eno1)[1860]: Program executed
> successfully.
> 
> On another ypserver in my network school (upgrade from mageia 4), ypserv
> keep is domainname after reboot but client users can not connect with ypbind
> (some clients run mageai4 others mageia 5). So I think something have
> changed or something is wrong in ypserv.
> 
> Looking for a bug I notice there is an error on clients with setsebool (does
> not find file) of selinux when starting ypbind. However selinux is not
> installed nor set.


Finally I think I found the problem. It is the authorized uid users (> 1000 in mga 5). I have modified in the makefile minuid and mingid parameters (/var/yp/makefile) in this way :
Commented script and replace bye number of the uid and gid authorized in previous Mageai.

.........line 39

# MINGID is the lowest gid that will be included in the group maps.
#MINUID=$(shell TMP=`egrep '^UID_MIN\s+[0-9]+' /etc/login.defs|sed -e 's/^UID_MIN//'` ; [ -n "$$TMP" ] && echo "$$TMP" || echo 1000)
#MINGID=$(shell TMP=`egrep '^GID_MIN\s+[0-9]+' /etc/login.defs|sed -e 's/^GID_MIN//'` ; [ -n "$$TMP" ] && echo "$$TMP" || echo 1000)
MINUID=500
MINGID=500

...............
Then compile tables with make in /var/yp 

So user id < 1000 can bind.

I think you have to update this part of makefile for users id between 500 (mageia 4) and 1000 (mageia 5).
Florian Hubold 2015-09-28 02:35:01 CEST

CC: (none) => doktor5000

Comment 15 Samuel Verschelde 2016-10-10 18:08:54 CEST
Joshua, it looks like the last comment is giving the solution. Will you fix the package?

Keywords: (none) => PATCH


Note You need to log in before you can comment on or make changes to this bug.