Perhaps you could use the heartbeat.config.monitors setting and point it to a directory? Then, in that directory, you could have one monitor configuration file per host? You could probably use a script to generate the 100 monitor configuration files from your file containing the 100 hosts?
Yes, I read this and make this default config into my cluster, but when it's related to a lots of machine that I need to monitor the icmp, I should put into the hosts in the file or create a lots of file, isn't it?
What I mean is like bash script that can use include file inside the monitor.d/icmp.yml that point the hosts, instead of hosts, we can use include file that contain the hundreds of hosts with the same configuration from the master config.
BTW, thanks for the answer, I still figure out how to sort this off.
@fadjar there aren't any performance concerns on our end if you generate a large icmp.yml file that contains hundreds of entries. Can you help us better understand why you'd avoid doing so?
I realized this related to the resources, so, I need to control the time.
The situation like this:
I have 1500 client POS entire country, since the application using centralized web apps, most of the time the remote POS always complain about speed.
I already add netflow in the router to check the response time and eggress/ingress kbps.
You know the user don't want to know if their network used for another purpose, just said slow response of the apps.... Hehehheee
So, I need the evidence for the response time of the entire POSes
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.