RabbitMQ internals monitoring by Zabbix

Continuation of extending zabbix-agent to monitor internals of applications. Now it's a RabbitMQ turn:


What's supported:

• File descriptors, Memory, Sockets watermarks monitoring
• Low level discovery of vhosts/queues
• Monitoring for messages, unack, consumers per queue
• Triggers for important counters
• Data sent in chunks, not one by one, using zabbix traps

Installation:

1. Save this as  /etc/zabbix/zabbix_agentd.conf.d/rabbitmq.conf
   

   # rabbitmq[server,uptime] will trigger trap sending
   # run '/etc/zabbix/rabbitmq.py server uptime debug' - to debug trap sending
   UserParameter=rabbitmq[*],/etc/zabbix/rabbitmq.py $1 $2
   

   
   
2. Which will call this data-getter
   /etc/zabbix/rabbitmq.py
   
   
3. Import template
   template_app_rabbitmq.xml 

Notes:

• Queues statistic is per-cluster and not per-server. This mean that having cluster of two nodes, both of them will have equal graphs for queues. So, there is line 31 in rabbitmq.py which states:
  

  if not socket.gethostname().endswith("-1"): return []  #only discover queues on first node

  We have naming convention like node-1, node-2 - and discovering queues only on first node. Remove/modify this line if it is not your case
• Data about queues is sent as zabbix traps to lower bandwidth and system load.
  To troubleshoot zabbix_sender run script manually as:
  

  /etc/zabbix/rabbitmq.py server uptime debug