Ceilometer is the code name of the OpenStack telemetry service. Ceilometer is responsible for metering in an OpenStack cluster. This service captures data about events, notifications, and measurements throughout the cluster. This data can be used for multiple purposes, including billing, monitoring, and autoscaling. Ceilometer works in conjunction with other OpenStack services in order to provide its functionality. In this chapter, we will explore the following topics:
A working Ceilometer installation will have several processes running, some of which are as follows:
ceilometer-api
ceilometer-collector
ceilometer-polling
ceilometer-agent-notification
When troubleshooting Ceilometer, an effective place to start is ensuring that each of the preceding processes run. As we've seen before, we can verify this by running either ps -aux
or pgrep
on Ceilometer. Take the following command, for example:
Alternatively, you can use the pgrep
command to retrieve similar information. An example is included in the following screenshot:
If any of these processes are not running as expected, you can search the log files for possible clues. The log files for Ceilometer are typically stored at /var/log/ceilometer
, but you can confirm the log file location by checking the Ceilometer configuration file for the value of the log_dir
variable. If the log files are empty or do not provide any clues, you can attempt to start the services manually, which will cause the output to be printed to the console as the process starts up. For example, to start ceilometer-api
manually, you would use a command similar to the following:
sudo -u ceilometer ceilometer-api --config-file=/etc/ceilometer/ceilometer.conf --log-file=/var/log/ceilometer/ceilometer-api.log
If there are errors during startup, examine the output carefully for clues pointing to the root cause.