infrastructure

Infrastructure

Please make sure all the prerequisites are met before proceeding.

A detailed tutorial for the cloud infrastructure setup is provided in the wiki.

To set up the credentials follow this guide, but additionally grant the following roles:

• roles/resources.editor (Editor)
• roles/storage.admin (Storage Admin)
• roles/bigquery.admin (BigQuery Admin)
• roles/logging.configWriter (Logs Configuration Writer)
• roles/resourcemanager.projectIamAdmin (Project IAM Admin)
• roles/serviceusage.serviceUsageAdmin (Service Usage Admin)

A short overview of the commands to set up the infrastructure and configure resources:

# Initialize
terraform init
terraform plan --out tfplan
terraform apply tfplan

terraform apply --var-file = experiment-3-syncmesh.tfvars

# Find one of the IPs and connect to the instance:
ssh -o StrictHostKeyChecking=no -L 8080 :ip : 8080 username@ip

# Follow Startup Script Log
sudo journalctl -u google-startup-scripts.service -f | grep startup-script
# See Startup Script Log
sudo journalctl -u google-startup-scripts.service | cut -d "]" -f2- | grep startup-script

# Start test manually
SLEEP_TIME=2 PRE_TIME=0 REPETITIONS=10 bash test.sh

gcloud auth activate-service-account terraform@dspj-315716.iam.gserviceaccount.com --key-file = "credentials.json"
gcloud config set project dspj-315716
gcloud compute instances get-serial-port-output  experiment-baseline-with-latency-3-test-orchestrator
# For Better display on smaller screens use
gcloud compute instances get-serial-port-output  experiment-distributed-gundb-with-latency-3-test-orchestrator | cut -d "]" -f2- | grep startup-script

# Destroy
terraform destroy

Experiments

Run the terraform the Terraform script for either baseline, syncmesh or advanced-mongo scenario:

terraform apply -var-file ./experiment-3-baseline.tfvars