|
||
---|---|---|
client | ||
config | ||
filters | ||
storage | ||
templates | ||
vendor | ||
.artifact_files | ||
.gitignore | ||
apiary.apib | ||
Dockerfile | ||
Gopkg.lock | ||
Gopkg.toml | ||
History.md | ||
LICENSE | ||
logResponseWriter.go | ||
main.go | ||
README.md | ||
strings.go | ||
structs.go | ||
web_handlers.go | ||
welcome_runner.go |
Luzifer / mondash
MonDash is a service for everyone having to display monitoring results to people who have not the time or knowledge to get familar with Nagios / Icinga or similar monitoring systems. Therefore MonDash provides a simple API to submit monitoring results and a simple dashboard to view those results.
Hosted
There is an instance of MonDash running on mondash.org you can use for free. This means you can just head over there, create your own dashboard with one click and start to push your own metrics to your dashboard within 5 minutes. No registration, no fees, just your dashboard and you.
Installation
However maybe you want to use MonDash for data you don't like to have public visible on the internet. As it is open source you also can host your own instance: The most simple way to install your own instance is to download a binary distribution on gobuild.luzifer.io.
This archive will contain the binary you need to run your own instance and the template used for the display. If you want just edit the template, restart the daemon and you're done customizing MonDash. If you do so please do me one small favor: Include a hint to this repository / my instance.
To start MonDash you will need to make sure you configured your instance correctly:
# mondash -h
Usage of mondash:
--api-token="": API Token used for the /welcome dashboard (you can choose your own)
--baseurl="": The Base-URL the application is running on for example https://mondash.org
--fileDirectory="./": Directory to use for plain text storage
--listen=":3000": Address to listen on
--s3Bucket="": Bucket to use for S3 storage
--storage="s3": Storage engine to use (s3, file)
- If you want to store the data in S3:
- Set AWS environment variables (
AWS_ACCESS_KEY_ID
,AWS_SECRET_ACCESS_KEY
,AWS_DEFAULT_REGION
) - Specify
--storage=s3
and--s3Bucket=<yourbucket>
- If you want to store the data in local file system:
- Ensure the data directory is writable
- Specify
--storage=file
and--fileDirectory=<yourdirectory>
In all cases you need to specify --api-token
with a token containing more than 10 characters and --baseurl
with the base-URL of your instance.
Docker
To launch it, just replace the variables in following command and start the container:
docker run \
-e AWS_ACCESS_KEY_ID=myaccesskeyid \
-e AWS_SECRET_ACCESS_KEY=mysecretaccesskey \
-e S3Bucket=mybucketname \
-e BASE_URL=http://mondash.org \
-e API_TOKEN=yourownrandomtoken \
-p 80:3000 \
luzifer/mondash
Security
Just some words regarding security: MonDash was designed to be an open platform for creating dashboards without any hazzle. You just open a dashboard, send some data to it and you're already done. No need to think about OAuth or other authentication mechanisms.
The downpath of that concept is of course everyone can access every dashboard and see every data placed on it. So please don't use the public instances for private and/or secret data. You can just set up your own instance within 5 minutes (okay maybe 10 minutes if you want to do it right) and you can ensure that this instance is hidden from the internet.