SONM is a powerful distributed worldwide system for general-purpose computing, implemented as a fog computing structure. Consumers of computing power in SONM get more cost-efficient solutions than cloud services., more information you can get at docs.sonm.com and sonm.com.
CockroachDB is SQL database for global cloud services. It's postgesql compatible. More information
This script manages to run CockroachDB distributed on SONM suppliers.
First of all you need to install sonmnode and sonmcli, you can do this by following this guide
Works only on Linux and MacOS, BSD systems not tested but may work.
Make sure you have installed all of the following prerequisites on your machine:
git clone https://github.com/akme/sonm-cockroach
Before deployment you need to set parameters of cluster in config.sh
When all parameters are set, run:
./rds.sh watch
This will create number of orders equal to numberofnodes you've set, when any order will become a deal, script will start task on it, each next task (next deal) will join cluster created by previous tasks.
When all orders created, all deals are set and have tasks running, it will watch for deals, if deal drops then it creates new orders, wait for deal and start task on it.
$ ./rds.sh watch
2018-08-17 23:49:21 Creating 3 order(s)
ID = 163493
ID = 163494
ID = 163496
2018-08-17 23:51:30 All set, waiting for deals
2018-08-17 23:51:30 watching cluster
2018-08-17 23:51:46 Starting task on deal 4090
Task ID: 749e08b4-bf1e-4601-99ed-640fdf1d9e33
Endpoint: 26257/tcp: 85.119.150.185:26257
Endpoint: 26257/tcp: 172.17.0.1:26257
Endpoint: 8080/tcp: 85.119.150.185:8080
Endpoint: 8080/tcp: 172.17.0.1:8080
Network: sonmbcfa4942
2018-08-17 23:51:58 Starting task on deal 4091
Task ID: 62d6d24d-ae2e-409d-8ac7-4ad983b08991
Endpoint: 8080/tcp: 185.144.156.200:8080
Endpoint: 8080/tcp: 172.17.0.1:8080
Endpoint: 26257/tcp: 185.144.156.200:26257
Endpoint: 26257/tcp: 172.17.0.1:26257
Network: sonmc44f619d
2018-08-17 23:52:13 Starting task on deal 4092
Task ID: 4e48b4e3-8d13-4a44-8d80-dcaf6ce5c04a
Endpoint: 26257/tcp: 95.141.193.156:26257
Endpoint: 26257/tcp: 172.17.0.1:26257
Endpoint: 26257/tcp: 172.18.0.1:26257
Endpoint: 8080/tcp: 95.141.193.156:8080
Endpoint: 8080/tcp: 172.17.0.1:8080
Endpoint: 8080/tcp: 172.18.0.1:8080
Network: sonm0e853006
After starting a task you will get an IP and port to access.
You can view cluster dashboard at http://
To get into SQL you can run:
./cockroach sql --certs-dir=certs --host <any node ip>
In case you want to access cluster with Postgres protocol you need to create user:
./rds.sh createuser <username>
It will ask for a password and will show you how to connect with psql, like this:
psql -h <any node ip> -p 26257 -U <username> --set=sslmode=require
You can get all cluster nodes IP addresses:
./rds.sh getips
By default cluster runs with replication factor 3, that means any data has 3 copies, but for such dynamic environment as fog, you may want to increase this number:
$ ./rds.sh setreplica 5
Cluster needs some time to make additional copies of data, so be patient to wait until sync ends.
When you want to stop using and destroy cluster, you just need to close all deals and cancel orders:
./rds.sh cancelorders
./rds.sh closedeals
It will stop all tasks and cleanup all created deals and orders.
This project is licensed under the MIT License - see the LICENSE file for details