Distibuted traffic

[stage]

This is more a question then a request for help.

I’m testing distributed monitoring.1 master and 2 node in all I’m testing 1vmware, 2 win2008, 1 netapp and 1 firewall. The traffic to the master node is 9.0528 message per second (so far the test). The goal is to monitor about 150 servers, 10 netapps, 15 VMware, a hand full of firewalls and nr of switches. So when I project the amount of 9 messages to the amount I want to do, there won’t be any room for other network traffic.

Distributed monitoring servers send messages to each other, when the test is don the result is send to the master distributed server. Yes?
How can I become that these message are as small as possible, and not more then necessary

I won’t that it only sends a message when there is a change in the triggers. The data of graphs, discovery, enz.. is from no impotents for the master distributed server. When I want to see the graph, I surf to de distributed server and not to the master distributed server

This to become a minimum of traffic and still can see all of the distributed nodes in the dashboard and the master distributed server can react with actions (mail or sms).

I’m waiting for your reaction,

Greets,
Dominic

[NOB]

Hi dominic

This is more a question then a request for help.

I’m testing distributed monitoring.1 master and 2 node in all I’m testing 1vmware, 2 win2008, 1 netapp and 1 firewall. The traffic to the master node is 9.0528 message per second (so far the test). The goal is to monitor about 150 servers, 10 netapps, 15 VMware, a hand full of firewalls and nr of switches. So when I project the amount of 9 messages to the amount I want to do, there won’t be any room for other network traffic.

Distributed monitoring servers send messages to each other, when the test is don the result is send to the master distributed server. Yes?
How can I become that these message are as small as possible, and not more then necessary

I won’t that it only sends a message when there is a change in the triggers. The data of graphs, discovery, enz.. is from no impotents for the master distributed server. When I want to see the graph, I surf to de distributed server and not to the master distributed server

This to become a minimum of traffic and still can see all of the distributed nodes in the dashboard and the master distributed server can react with actions (mail or sms).

I’m waiting for your reaction,

Greets,
Dominic

There are the following options for the zabbix_server

1. NodeNoEvents
2. NodeNoHistory

I think, the latter is what you want to enable.

HTH,

Norbert.

[NOB]

Hi stage

you shouldn't open a new thread for just asking what it is and does.
I am sorry, but I was in a hurry answering your question.

It does what it says, I guess:

• NodeNoEvents will send no events, just the data
• NodeNoHistory will send no historical data, just events

You configure it in the remote ZABBIX-Server (connected to a master).

HTH,

Norbert.