Fabric8 Documentation


Fabric8 is built for distributed deployments: distributed across container, VMs & datacentres. Each container logs & gathers metrics which is great, but viewing those logs & metrics across a distributed deployment isn't going to be easy. So Fabric8 provides a consolidated view on logs & metrics collected in your fabric, making it easy to know exactly what's going on in your entire deployment.

Logs & Metrics

Insight collects logs & metrics separately, allowing you to choose what you want to collect & where you want to store your data. Out of the box, Fabric8 Insight collects a reasonable amount of data so you have to think about how you're going to deploy Insight. Currently, Insight stores log data in Elasticsearch & metrics data can be stored in either Elasticsearch, Cassandra (through RHQ Metrics), or InfluxDB (only available when using Fabric8 with Docker). This can all be run on one node (see installation below), but in a production deployment you are recommended to deploy a separate cluster of nodes to collect your data. This cluster can all be managed through Fabric8 profiles.


Installing Insight is very simple: just as with anything else in Fabric8, it's all done by assigning profiles. To create a data node for Elasticsearch, just assign the insight-elasticsearch.datastore profile to an existing container or create a new one. If you're doing this all on one container you can simply do:

container-add-profile root insight-elasticsearch.datastore

Or if you want to create a separate cluster of 3 Elasticsearch nodes you can simply run:

container-create-child --profile insight-elasticsearch.datastore root ds 3

Elasticsearch will use the Fabric8 Zookeeper registry for discovery.

So now you have somewhere to store your data, you now need to tell your containers to start logging to Elasticsearch. Just add the insight-logs.elasticsearch profile to any container that you want to send its logs to Elasticsearch:

container-add-profile root insight-logs.elasticsearch

Now that you have logs being collected in Elasticsearch, you need something to view & filter them. This is provided by the insight-console profile:

container-add-profile root insight-console

Hawtio will then have a new perspective available: Fabric8. This will have a tab called Logs which provides a Kibana dashboard for you to search all logs across your fabric.

The Insight Console

The Insight Console uses the Hawtio Kibana plugin. This provides a very flexible & powerful way to inspect your logs & find what you’re looking for. The Log dashboard looks like this:

Viewing Insight Logs

Each log entry can be expanded to view full details of the log event & stacktraces are formatted nicely, including linking to source code for easy fault finding:

Expanding Log Events