About the collection of system monitoring data
All PingDirectory servers have the capability to monitor the health of the server and host system they run on for diagnostic review and troubleshooting.
This topic applies only to the PingDirectory server. |
Initially, the servers do not collect any performance data until they are prepared for monitoring by a Metrics server using the monitored-servers add-servers
tool or an administrator enables system health data collection for real-time inspection and querying. At a high level, all of the important server and machine metrics that can be monitored are available in the cn=monitor
backend.
The Stats Collector plugin relies exclusively on entries in the cn=monitor
backend to sample data using LDAP queries. The Stats Collector plugin is the primary driver of performance data collection for LDAP, server response, replication, local Java Runtime Environment (JRE) databases, and host system machine metrics. Stats Collector configuration determines the sample and collection intervals, granularity of data (basic, extended, verbose), types of host system collection (CPU, disk, network) and the type of data aggregation that occurs for LDAP application statistics. The Stats Collector plugin is configured with the dsconfig
tool and collects data using LDAP queries.
For example, the --server-info:extended
option includes collection for the following:
-
CPU
-
Java virtual machine (JVM) memory
-
Memory
-
Disk information
-
Network information
Utilization metrics are gathered through externally invoked OS commands, such as iostat
and netstat
, using platform-specific arguments and version-specific output parsing.
Enabling the Host System monitor provider automatically gathers CPU and memory utilization but only optionally gathers disk and network information. Disk and network interfaces are enumerated in the configuration by device names, such as eth0
or lo
, and by disk device names, such as sd1, sdab, sda2,
scsi0
.