Change of IP's will result in crash of mmmd_mon upon startup

[GoogleCodeExporter]

To recreate:

Setup a stable working configuration.  Stop mmmd_mon and add/remove IP
addresses from the mmm_mon.conf file.  Restart mmmd_mon and check status:

# ./mmm_control show
Config file: mmm_mon.conf

WARNING!!! DAEMON IS NOT RUNNING. INFORMATION MAY NOT BE ACTUAL!!!

Servers status:
[...]

Original issue reported on code.google.com by bn-s...@nesbitt.org on 11 Apr 2007 at 6:58

[GoogleCodeExporter]

Here is the output of the startup after changing:

# Mysql Writer role
role writer
    mode exclusive
    servers test01, test02
    ip 10.0.1.15, 10.0.1.16

to:

# Mysql Writer role
role writer
    mode exclusive
    servers test01, test02
    ip 10.0.1.15

# /etc/init.d/mmm_mon start
Daemon bin: ''
Daemon pid: ''
Starting MMM Monitor daemon: /usr/local/mmm/sbin/mmmd_mon
Reading config file: 'mmm_mon.conf'
$VAR1 = {
          'test03' => {
                        'roles' => [
                                     'reader(10.0.1.11;)'
                                   ],
                        'version' => '0',
                        'state' => 'ONLINE'
                      },
          'test02' => {
                        'roles' => [
                                     'reader(10.0.1.12;)'
                                   ],
                        'version' => '0',
                        'state' => 'ONLINE'
                      },
          'test01' => {
                        'roles' => [
                                     'writer(10.0.1.16;)',
                                     'writer(10.0.1.15;)'
                                   ],
                        'version' => '0',
                        'state' => 'ONLINE'
                      }
        };
Role: 'reader(10.0.1.11;)'
Adding role: 'reader' with ip '10.0.1.11'
Role: 'reader(10.0.1.12;)'
Adding role: 'reader' with ip '10.0.1.12'
Role: 'writer(10.0.1.16;)'
Adding role: 'writer' with ip '10.0.1.16'
Invalid value for shared scalar at /usr/local/mmm/lib/config.pm line 274.
Ok

Original comment by bn-s...@nesbitt.org on 11 Apr 2007 at 7:29

[GoogleCodeExporter]

The following procedure might solve the problem:
1. stop mmmd_mon
2. stop mmmd_agent on related machines
3. restart mmmd_agent on related machines
4. restart mmmd_mon

Original comment by sms2...@gmail.com on 31 Jul 2008 at 9:15

[GoogleCodeExporter]

To solve the problem you have to delete the file mmmd.status on the monitoring 
host.
Though this should be fixed in mmm.

Original comment by m...@pascalhofmann.de on 15 Oct 2008 at 3:36

[GoogleCodeExporter]

Fixed in rev# 112.

Original comment by m...@pascalhofmann.de on 15 Oct 2008 at 4:13

• Changed state: Fixed