https://github.com/dropbox/grallama-panel
GraLLAMA panel for LLAMA data
https://github.com/dropbox/grallama-panel
latency llama loss matrix monitoring network telemetry
Last synced: 3 months ago
JSON representation
GraLLAMA panel for LLAMA data
- Host: GitHub
- URL: https://github.com/dropbox/grallama-panel
- Owner: dropbox
- License: other
- Created: 2017-02-24T22:57:05.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2025-03-28T18:00:41.000Z (about 1 year ago)
- Last Synced: 2026-03-04T18:58:33.556Z (3 months ago)
- Topics: latency, llama, loss, matrix, monitoring, network, telemetry
- Language: JavaScript
- Homepage:
- Size: 4.06 MB
- Stars: 16
- Watchers: 13
- Forks: 7
- Open Issues: 9
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Grafana GraLLAMA Panel
GraLLAMA is a panel plugin for Grafana which visualizes the matrix-like data
collected by the [LLAMA](https://github.com/dropbox/llama) project. See the
LLAMA project on GitHub for more information:
https://github.com/dropbox/llama
## Features
* Displays loss percentages between sources and destinations.
* Sources/Destinations are dynamically pulled from timeseries tags.
* Changes cell color based on higher percentages.
## Compatibility
This panel should work with the following datasources:
[InfluxDB](https://grafana.net/plugins/influxdb)
## Installation
Use the new grafana-cli tool to install grallama-panel from the commandline:
```
grafana-cli plugins install dropbox-grallama-panel
```
The plugin will be installed into your grafana plugins directory; the default
is /var/lib/grafana/plugins if you installed the grafana package.
More instructions on the cli tool can be found
[here](http://docs.grafana.org/v3.0/plugins/installation/).
You need the lastest grafana build for Grafana 3.0 to enable plugin support.
You can get it here : http://grafana.org/download/builds.html
## Setup
After installing the panel plugin follow these instructions to visualize
the latest loss data from your running LLAMA deployment.
0. Setup a datasource for InfluxDB where LLAMA data is stored.
1. Add a new panel to a row in Grafana and select GraLLAMA.
2. Open the query editor and select InfluxDB and the measurement `loss`.
3. Select `mean()` as an aggregation (or others if you see fit).
4. Select `Group by` using exactly two tags; one of them will be the source
tag and the other will be the destination tag. Also, use `fill(null)`.
5. Type into the `Alias by` field `-` (the hypen is
important). If your source tag was `foo` and destination tag was `bar` then
the field would look like this: `$tag_foo-$tag_bar`