`rmdef` command do not help clean up the recousrces that related to
the node when removing the node. this is a technical debt. This patch
is only a work around to add -C|--cleanup option to help remove the
entries in goconserver.
On diskless SN, the nodes file of goconserver is missing when
reinstalling. This patch add a column in nodehm table to help
recover the console information on service node.
```
[root@c910f03c09k04 ~]# makegocons -q
NODE SERVER STATE
c910f03c09k05 c910f03c09k04.pok.stglabs.ibm.com error
node-7912ac1-06vag36 c910f03c09k04.pok.stglabs.ibm.com error
NODE SERVER STATE
c910f03c09k04 c910f03c09k03.pok.stglabs.ibm.com connected
[root@c910f03c09k04 ~]# rm /var/lib/goconserver/*
rm: remove regular file ‘/var/lib/goconserver/nodes.json’? y
rm: remove regular file ‘/var/lib/goconserver/nodes.json.bak’? y
[root@c910f03c09k04 ~]# service xcatd restart
Restarting xcatd (via systemctl): [ OK ]
[root@c910f03c09k04 ~]# makegocons -q
NODE SERVER STATE
c910f03c09k05 c910f03c09k04.pok.stglabs.ibm.com error
node-7912ac1-06vag36 c910f03c09k04.pok.stglabs.ibm.com error
NODE SERVER STATE
c910f03c09k04 c910f03c09k03.pok.stglabs.ibm.com connected
```
Add `-q|--query [noderange]` for makegocons to list the console
status and the host server. If node is not registered in
goconserver, the status will be `unregistered`.
If console releated attribute is not defined for the specified
node but it is included in the noderange, error message for that
node will be printed at first.
- clear logs in the key components during provisioning (not only debug mode, but debug mode will have more logs)
- now only cover rhel7 stateful and stateless
- make more syslogs into cluster.log (tag = xcat or xcat.*)
- the changes will not impact other distro
Previously, it is impossible to know if the goconerver is started
by xcat. This patch add a marker in /etc/goconserver/server.conf
to help check the status. In addition, use server-cred.pem instead
of server-key.pem as it is not exist on SN.
This patch also control the service status of goconserver and conserver
when restarting the xcatd service on service node.
```
[root@sn02 ~]# chdef sn02 setupconserver=1
1 object definitions have been created or modified.
[root@sn02 ~]# service xcatd restart
Restarting xcatd (via systemctl): [ OK ]
[root@sn02 ~]# ps -ef | grep conserver
root 27679 1 0 02:26 ? 00:00:00 /usr/sbin/conserver -o -O1 -d
root 27680 27679 0 02:26 ? 00:00:00 /usr/sbin/conserver -o -O1 -d
root 27756 26188 0 02:26 pts/1 00:00:00 grep --color=auto conserver
[root@sn02 ~]# chdef sn02 setupconserver=2
1 object definitions have been created or modified.
[root@sn02 ~]# service xcatd restart
Restarting xcatd (via systemctl): [ OK ]
[root@sn02 ~]#
[root@sn02 ~]#
[root@sn02 ~]#
[root@sn02 ~]# ps -ef | grep conserver
root 27885 1 0 02:26 ? 00:00:00 /usr/bin/goconserver
root 27986 26188 0 02:33 pts/1 00:00:00 grep --color=auto conserver
```
This patch refactor the interface to get xcatmaster and tftpserver
attributes in the subvars subroutine to improve the performance.
partial-issue: #3958
* Fix#3428, set xcatmaster to facing IP in priority when the node is mananged by service node.
* Always using my facing IP first than site.master
* remove non-used variable sitemaster
In some scenarios, an IPMI instance could be called to handle
data earlier than it should. Ensure the state of the object is
in adequate shape to react as soon as possible.
* fix issue [DEV] nodes provisioned with redhat osimages created by xCAT 2.12.3 or earlier missed yum repo to MN in xCAT 2.12.4 or later #2856
* fix issue [DEV] nodes provisioned with redhat osimages created by xCAT 2.12.3 or earlier missed yum repo to MN in xCAT 2.12.4 or later #2856
