DM Replication Delay

[pedroindio]

Hi,

I have a DM implementation with more than 50 Nodes.
I have more than 1250 Hosts beeing monitored and all the alerts are to be sent by Master Node.

The problem is that i have been notice that the alerts are sent with a large delay (sometimes more than an hour).
I have been making some debug and i notice that data is not being replicated to Master Node as it should be... at least faster.

Since the alerts are to be sent my Master Node this is a big problem.

I think the problem can be related with the housekeeping (Master Node and/or Chield Nodes) but i am not sure.

I would like to solve this issue because this solution will grow for at least 100 Nodes.

any idea about why i am having this problem and how can i fix it?

Another note: The Zabbix Server Frontend sometimes indicates that "Zabbix Server is not running: ..." but it is and i dont have any triggers related with the zabbix service or database service.


My configurations, at the moment, are:

[Node Master]:

2 Servers:

• 1 for zabbix server:
  Hardware: VMWare Host with 1vCPU (2,133Ghz), 4GB RAM, HA
  Software: Zabbix-Server 2.0.5
  Configurations: 5 Hosts
  
• 1 DB server:
  Hardware: Physical HW Quad-Core CPU (3.6 GHz), 8GB RAM, HW RAID 1
  Software: MySQL SQL Server 5.5
  DB Configurations: innodb tables (no table partitions)
  

[Chield Nodes]

For each node,

1 Server:

• 1 Server: Hardware: VMWare Host with 1vCPU (3,1Ghz), 4GB RAM, HA
  Software: Zabbix-Server 2.0.5
  Configurations: 25+ Hosts
  DB Configurations: innodb tables (no table partitions)

thanks in advance,