This is FAQ about installation step.
Scouter has agent-server-client architecture. Scouter Agent collects monitoring data and sends it to Scouter Server. Scouter Server processes and accumulates received data and sends it to Scouter Client by response result. Scouter Client reads processed monitoring and performance data from Server, display them on the user's screen with various views and perspectives.
Agent[Linux] ----\
Agent[Tomcat] ------> Server ——-> Client
Agent[MariaDB] ----/
Currently there are Tomcat, Linux, and MariaDB agents. In the future more agents for more opensource software would be developed and included.
Scouter's three SW use TCP and UDP for their communication method. With configuring IP addresses and ports along to your environment, you can install three SW independently.
Scouter Agent uses both TCP and UDP. Agent reports its data via UDP to Scouter Server periodically. And uses TCP when reply Server's special requests.
UDP 6100
Agent ====> Server
TCP 6100
Agent is configured to use 127.0.0.1 for Server's IP address. To change this, modify java options on Agent program.
-Dsever_addr=192.168.0.1
-Dserver_udp_port=6100
-Dserver_tcp_port=6100
Also configurable to modify configuration properties file.
sever_addr=192.168.0.1
server_udp_port=6100
server_tcp_port=6100
If you want to another configuration file, uses -Dscouter.config= option.
Scouter Client uses TCP 6100 port to request monitoring data, same as Agent does.
Client ——-> Server
TCP 6100
TCP 6100 port are used by both Agent and Client. User should specify Server's endpoint information when Client program run at first.
Server Address: 192.168.0.1:6100
Scouter used logical names and hierarchy for its monitoring targets.
/host/tomcat/context
If you want to watch multiple Tomcat instances, you should give unique name of each on the same hierarchy. Names can be given by java options like,
-Dobj_name=MyTomcatInstName
or on scouter.config file,
obj_name=MyTomcatInstName
TBD