These FAQ address how to enable and use OpenStack notification logging.
What information is available in OpenStack notifications?
OpenStack services note events related to the changing states of cloud assets such as instances, images, volumes, and network resources. Examples of events include creation, updating, or deletion of an instance or the attaching of a volume to an instance.
How can these notifications be useful?
Notifications provide additional visibility into OpenStack activity and allow you to capture history. While the Dashboard or CLI provide a "live" view into the current state of your cloud, a log of notifications is a record of past events. This can be useful for troubleshooting, for example, tracking the lifespan of an instance or volume.
How can my organization receive OpenStack notifications?
The notification service is provided via a standard syslog stream, which can be consumed by a log processing system, such as Splunk, Logstash, or Greylog.
What do I need to do to enable logging of the notifications?
Submit a P4 request to Metacloud Support with the IP address or fully qualified domain name of a log host, as well as its connection port. It is recommended, though not required, that you use a log server that is not in the cloud, in case cloud service is disrupted.
Make sure your firewall access control list (ACL) allows incoming traffic from your Metacloud Control Planes (MCPs), which send out the events. If you do not have the IP addresses of the MCPs, ask Metacloud Support for them.
What protocols are supported for notification traffic?
Metacloud supports UDP. It does not support TCP because of the risk of congestion in the message queue if the log server becomes unavailable. This congestion could potentially disrupt MCP activity.
Will the enabling of notification logging disrupt my cloud operations?
Running instances are not affected. The configuration work requires a restart of the OpenStack services. The Dashboard and APIs are unavailable for approximately 20 minutes.