Updated Oct 15, 2008 by m...@pascalhofmann.de
Labels:	Phase-Deploy

ConfigurationExamples

Config file examples with a short description.

Introduction

Detailed description of this example configuration could be found in “Typical use-cases of mmm-deployment” document.

Master-master replication example

Let’s take a look at an example configuration file for such configuration. Every deployment of mmm requires N+1 installation where N is number of MySQL servers and 1 is monitoring server (can be combined with application server). Following examples would be named as mmm_mon.conf for monitoring server and mmm_agent.conf for database servers.

mmm_mon.conf:

#
# Master-Master Manager config (monitor)
#

# Debug mode
debug no

# Paths
pid_path /opt/mmm/var/mmmd.pid
status_path /opt/mmm/var/mmmd.status
bin_path /opt/mmm/bin

# Logging setup
log mydebug
    file /opt/mmm/var/mmm-debug.log
    level debug

log mytraps
    file /opt/mmm/var/mmm-traps.log
    level trap


# MMMD command socket tcp-port
bind_port 9988
agent_port 9989
monitor_ip 127.0.0.1

# Cluster interface
cluster_interface eth0

# Cluster hosts addresses and access params
host master1
    ip 192.168.1.8
    port 3306
    user rep_monitor
    password RepMonitor
    mode master
    peer master2

host master2
    ip 192.168.1.9
    port 3306
    user rep_monitor
    password RepMonitor
    mode master
    peer master1

#
# Define roles
#

active_master_role writer

# Mysql Reader role
role reader
    mode balanced
    servers master1, master2
    ip 192.168.1.100, 192.168.1.101

# Mysql Writer role
role writer
    mode exclusive
    servers master1, master2
    ip 192.168.1.200

#
# Checks parameters
#

# Ping checker
check ping
    check_period 1
    trap_period 5
    timeout 2

# Mysql checker
check mysql
    check_period 1
    trap_period  2
    timeout 2

# Mysql replication backlog checker
check rep_backlog
    check_period 5
    trap_period 10
    max_backlog 60
    timeout 2

# Mysql replication threads checker
check rep_threads
    check_period 1
    trap_period 5
    timeout 2

In this config file I want to highlight few sections:

Cluster hosts addresses and access params - this section describe IPs and credentials that will be used in all monitoring checks. So, it is required to have an active MySQL instance listening on these IPs and appropriate access restrictions set to allow monitoring server to MySQL instance.
Define roles - this section defines what roles we have in our cluster and what IPs will be used for each role. Since roles are moved among servers each server needs an IP (for the monitor to talk to it) and each role needs an IP.

mmm_agent.conf:

#
# Master-Master Manager config (agent)
#

# Debug mode
debug no

# Paths
pid_path /opt/mmm/var/mmmd_agent.pid
bin_path /opt/mmm/bin

# Logging setup
log mydebug
    file /opt/mmm/var/mmm-debug.log
    level debug

log mytraps
    file /opt/mmm/var/mmm-traps.log
    level trap

# MMMD command socket tcp-port and ip
bind_port 9989

# Cluster interface
cluster_interface eth0

# Define current server id
this master1
mode master

# For masters
peer master2

# Cluster hosts addresses and access params
host master1
    ip 192.168.1.8
    port 3306
    user rep_monitor
    password RepMonitor

host master2
    ip 192.168.1.9
    port 3306
    user rep_monitor
    password RepMonitor

All other config sections are pretty straightforward and described in detail in configuration files section of MMM guide.

Comment by vit.bartunek, Mar 04, 2008

Is it correct to have 192.168.1.201 in role writer section? I supposed that we just move 192.168.1.200 between servers..

# Mysql Writer role role writer

mode exclusive servers master1, master2 ip 192.168.1.200, 192.168.1.201

Comment by fernando.castano, Oct 13, 2008

I have the same question. was this question ever resolved? it is from march (!!)

Comment by m...@pascalhofmann.de, Oct 15, 2008

It was an error. The following is correct:

role writer

mode exclusive servers master1, master2 ip 192.168.1.200

Comment by zhou.shq, Nov 16, 2008

log level is: trap, error, notice, warn, debug 'trap' = 0; 'error' = 10; 'notice' = 20; 'warn' = 30; 'debug' = 100;

Comment by guo...@rocketmail.com, Jan 29, 2009

I am experiencing a connection problem after moving writer role to another machine or setting one db is offline. I need wait for several minutes for a ready connection. Sometimes the waiting time is 7 minutes. To be clear,

db1(209.134.190.115): master/ONLINE. Roles: reader(209.134.190.119;), writer(209.134.190.120;) db2(209.134.190.116): master/ONLINE. Roles: reader(209.134.190.118;)

If I move the writer to db2, then connection to 209.134.190.120 cannot be established immediately, instead, my perl script is timed out. After that, the connection resumes.

How can I work around this?

Thank you,

Coway

Comment by fuzzygroove, Mar 24, 2009

This example is outdated. Make sure to look at the examples that come with the project source instead.