Deploy-Workload-API.md 1.34 KB
Newer Older
Daniel Seybold's avatar
Daniel Seybold committed
1 2 3 4 5 6 7 8 9 10
#### Install Workload-API 

In order to achieve comparable results to the ones presented in the paper a VM with preferable the following resources is required:

| property | required |
| --- | --- |
| image / operating system | Ubuntu 16.04 |
| CPU cores | 8 |
| memory | >= 4GB |
| disk | 20GB |
11
| software | Docker |
Daniel Seybold's avatar
Daniel Seybold committed
12 13 14 15 16 17 18 19 20 21
| open ports | 8181 |

**Task 1:** Go to your Cloud provider and create a VM of the aforementioned configuration.  
**Task 2:** As the Workload-API needs to publicly expose port `8181`, add a security group rule to the VM for port `8181` that allows in-going traffic.   
**Task 3:** Assign public IP to the VM. We denote this IP in the following as *WORKLOAD_API_IP*.   
**Task 4:** Log into the VM, and install Docker and Docker-Compose.  
**Task 4a:** Docker can be installed as described in: https://docs.docker.com/install/linux/docker-ce/ubuntu/.  

#### Start Workload-API 

22
**Task 1:** Get the public IP this  VM and export it via `export PUBLIC_IP=X.X.X.X`
Daniel Seybold's avatar
Daniel Seybold committed
23

24
**Task 2:** Get the public IP of the VM that hosts the Mowgli VM and export it via `export INFLUX_IP=X.X.X.X`
Daniel Seybold's avatar
Daniel Seybold committed
25

Daniel Seybold's avatar
Daniel Seybold committed
26
**Task 3:** Now run the Workload-API (provided as Docker container) with the following command `docker run -d -p 8181:8181 -e PUBLIC_IP=$PUBLIC_IP -e INFLUXDB_URL=$INFLUX_IP:8086 -v /tmp:/opt/results -v /tmp:/var/log omi-registry.e-technik.uni-ulm.de/mowgli/workload-api:0.1`
Daniel Seybold's avatar
Daniel Seybold committed
27