teleport/assets/monitoring
2017-05-28 15:37:21 -07:00
..
convert.py monitoring, docs and fix forward on conn close 2017-05-28 15:37:21 -07:00
docker-compose.yml instrument with monitoring tools, fixes #935 2017-05-13 18:32:10 -07:00
env.influxdb instrument with monitoring tools, fixes #935 2017-05-13 18:32:10 -07:00
gops.py monitoring, docs and fix forward on conn close 2017-05-28 15:37:21 -07:00
grafana-import.sh instrument with monitoring tools, fixes #935 2017-05-13 18:32:10 -07:00
health-dashboard.json monitoring, docs and fix forward on conn close 2017-05-28 15:37:21 -07:00
README.md monitoring, docs and fix forward on conn close 2017-05-28 15:37:21 -07:00
telegraf.conf fix bench goroutine leak and add more stats 2017-05-14 12:47:04 -07:00
teleport.service add some monitoring units 2017-05-13 19:27:38 -07:00
tick.service add some monitoring units 2017-05-13 19:27:38 -07:00

Metrics

Teleport exports Prometheus Metrics endpoints.

Start Teleport With Prometheus Endpoint

To start teleport with prometheus endpoint enabled:

teleport start --diag-addr=127.0.0.1:3434

To start monitoring stack, simply docker-compose up

Then go to http://localhost:3000

Username: admin
Password: admin

Grafana Dashboard

Modify and export grafana dashboard, then convert to the format that Grafana can auto import using this tool

python convert.py health-raw.json health-dashboard.json

Low level monitoring

Teleport adds gops as a low level debugging solution:

teleport start --gops --gops-addr=127.0.0.1:4321

Then to use gops:

go get github.com/google/gops
gops stack $(pidof teleport

Diffing goroutine dumps

We have a tool to give you idea of the difference between two teleport stack dumps, so we can see what's the overhead and difference to detect leaks:

gops stack $(pidof teleport) | python gops.py collect > /tmp/a
# take a second diff
gops stack $(pidof teleport) | python gops.py collect > /tmp/b
# compare two diffs
python gops.py diff /tmp/a /tmp/b