feat(infra): Z→Q migration — Ansible/Terraform/monitoring rename (PR 2/3)

[moscowchill]

Summary

• Renames the QRL execution-client binary gzond → gqrl and Ansible role directory to match the upstream project rename (go-zond → go-qrl).
• Renames all Ansible variables gzond_* → gqrl_* (defaults, playbooks, inventory, group_vars), cross-role references (qrysm-beacon, security, monitoring), and the systemd service template.
• Renames Terraform variables zond_rpc_url → qrl_rpc_url, env ZOND_RPC_URL → QRL_RPC_URL across modules (validator-node, backup-node, monitoring-server, testnet env).
• Renames Prometheus scrape job gzond → gqrl, alert rules GzondDown/LowPeers/Syncing → Gqrl*, Grafana dashboard queries, Alertmanager inhibit rules.
• Updates infrastructure/README.md, infrastructure/docs/ARCHITECTURE.md, monitoring/README.md, and inline comments.

Scope note

Runbooks and standalone documentation (infrastructure/docs/runbooks/*, DEPLOYMENT.md, validator-integration.md, top-level README.md, docs/architecture.md) are intentionally out of scope and will be covered by PR 3.

Files changed (33)

• 1 Ansible role dir renamed + service template renamed
• 14 Ansible files updated
• 8 Terraform files updated
• 3 shell scripts updated
• 5 monitoring config/dashboard files updated
• 2 top-level README/ARCHITECTURE docs adjacent to infra

Test plan

• Ansible dry-run on a disposable host: ansible-playbook -i inventory.ini.template playbooks/deploy-node.yml --check --diff
• terraform -chdir=infrastructure/terraform validate + terraform plan -var-file=environments/testnet/terraform.tfvars
• promtool check rules monitoring/prometheus/rules/validator-alerts.yml
• promtool check config monitoring/prometheus/prometheus.yml
• Grafana dashboard JSON imports cleanly (visual)
• Grep for residual gzond, Zond, ZOND_ in changed dirs (excluding runbooks/docs) — should be empty

[gemini-code-assist]

Code Review

This pull request performs a comprehensive rename of the execution client from gzond to gqrl across the entire infrastructure codebase, including Ansible roles, Terraform modules, monitoring, and documentation. Feedback was provided to improve the gqrl Ansible role by ensuring the build task correctly handles binary updates, refining the sync status display for clarity, and utilizing the defined log level variable in the systemd service template.

[gemini-code-assist]

The creates argument in this task will prevent the gqrl binary from being updated when the source code changes. Since the binary already exists at gqrl_binary_path after the first deployment, Ansible will skip this task even if gqrl_clone.changed is true. Additionally, the when condition gqrl_clone.changed or not ansible_check_mode is problematic as it will attempt to run the build on every normal execution (since not ansible_check_mode is true), which is inefficient.

Consider removing creates and using a more robust when condition that checks if the source changed or if the binary is missing.

- name: Build gqrl from source
  shell: |
    cd /opt/quantapool/gqrl/source
    make gqrl
    cp build/bin/gqrl {{ gqrl_binary_path }}
  when: gqrl_clone.changed or not ansible_check_mode
  notify: restart gqrl

[gemini-code-assist]

The sync status message is counter-intuitive when the node is fully synced. The eth_syncing RPC method returns false when the node is synced, which results in the message gqrl sync status: False. It would be clearer to explicitly state 'Synced' or 'Syncing'.

    msg: "gqrl sync status: {{ 'Synced' if sync_status.json.result == false else 'Syncing' }}"

[gemini-code-assist]

The verbosity level is hardcoded to 3, which ignores the gqrl_log_level variable defined in the role's defaults (infrastructure/ansible/roles/gqrl/defaults/main.yml). It is better to use the variable to allow environment-specific overrides.

  --verbosity {{ gqrl_verbosity | default(3) }}