This will download logs and metrics from the cluster to your local machine.
cd deploy
ansible-playbook --user ec2-user --inventory `which terraform-inventory` collect_logs_and_metrics.yaml
Example output:
ls ../../data/pravega_logs/pravega_logs_20191004T034401
benchmark-worker-10.0.0.119.log.gz
benchmark-worker-10.0.0.13.log.gz
benchmark-worker-10.0.0.42.log.gz
benchmark-worker-10.0.0.82.log.gz
bookkeeper-10.0.0.145.log.gz
bookkeeper-10.0.0.186.log.gz
bookkeeper-10.0.0.247.log.gz
influxdb.tgz
pravega-controller-10.0.0.242.log.gz
pravega-segmentstore-10.0.0.121.log.gz
pravega-segmentstore-10.0.0.253.log.gz
pravega-segmentstore-10.0.0.80.log.gz
prometheus.tgz
zookeeper-10.0.0.185.log.gz
zookeeper-10.0.0.202.log.gz
zookeeper-10.0.0.219.log.gz
This will run local instances of InfluxDB, Prometheus, and Grafana loaded with previously collected metrics.
cd deploy
open_saved_metrics/open_saved_metrics.sh ../../data/pravega_logs/pravega_logs_20191004T034401
Open Grafana at http://localhost:3000. Login using user name "admin" and any password.
Configure Grafana with the following data sources:
- Prometheus
- Name: Prometheus
- HTTP URL: http://prometheus:9090
- InfluxDB
- Name: pravega-influxdb
- HTTP URL: http://influxdb:8086
- InfluxDB Details Database: pravega
Load dashboards from deploy/templates/dashboards.