The TICK stack is a very powerful architecture for managing time-series data. It can be used for everything from monitoring server infrastructure to process IoT data. Unfortunately hosted InfluxCloud starts at $99 a month which might be too much when considering it for experiments or prototypes.

Thankfully Matt pointed me to interesting Docker hosting: Hyper.sh. One of the reasons I like Hyper is quicker setup and going with "Heroku for Docker" approach making it perfect for proof on concepts or small deployments. It does support Docker Compose file format so basing on InfluxData Compose file I created one which works on Hyper.sh:

👉🏻 Follow instructions in hyper-tick repository

What's the cost of this?

Here's the breakdown based on Hyper.sh pricing:

• $1 - FIP ("Floating" IP, basically a public IP)
• 2x $1.55 - two S2 containers (haproxy and influxdb crash with just 64MB of ram)
• 2x $1.03 - two S1 containers for chronograf and kapacitor
• 4x $0.1/GB - four images
• 3x $0.1/GB - three volumes attached to influxdb, chronograf and kapacitor (this might grow with amount of your data stored in InfluxDB)

Making it total of $6.86 per month. Just low enough to allow everyone start their adventure with TICK stack.

What can I do with this?

If you're not familiar with TICK stack, here's list of possible applications from InfluxData use cases:

• Custom DevOps Monitoring
• Real Time Analytics
• IoT and Sensor Data
• Cloud & OpenStack
• Anomaly Detection
• Messaging
• Personalization
• Equities Trading
• Municipal Infrastructure Management
• GPS Services
• Quantum Physics Research
• Point of Sale Systems
• Manufacturing & Home Automation
• Transportation & Material Logistics

What I'm currently using it for is server monitoring and alerting:

To report system stats you need to install Telegraf on your server and point it to your InfluxDB. Then it's just a matter of creating visualisations in Chronograf and alerts in Kapacitor.