solscope-validator-watcher 0.3.0

Standalone cron-based monitoring and alerting for Solana validators

solscope-validator-watcher

Cron-friendly Solana validator monitoring you can run on your own validator host. It reproduces the alerting that SolScope provided as a hosted service, with no server or database required: a config file, a state file for cooldowns, and a one-minute cron job.

It ships with a full-screen terminal UI (built on Textual) for managing multiple validators — each with its own watchers and notification channels — plus a non-interactive run-once command for cron.

Watchers

Watcher	What it checks
sfdp_version	Your node's version against the SFDP required minimum from api.solana.org. Detects whether the node runs Agave or Firedancer (by its version line) and compares against that client's minimum (agave_min_version or firedancer_min_version).
software_outdated	Your node's version against the latest stable Agave release on GitHub (anza-xyz/agave) within your node's current major version line (e.g. a v3.1.x node is compared to the newest v3.x.y).
delinquent	Whether your vote account is reported delinquent via getVoteAccounts.

Each watcher has its own cooldown_minutes so a per-minute cron won't spam you while a condition persists. Cooldown state is tracked in a JSON state file alongside the config.

Notification channels

Configure any combination under notifications:

• slack_webhooks — Slack incoming webhook URLs
• discord_webhooks — Discord webhook URLs
• webhooks — generic webhooks ({"text": "..."} POST body)
• ntfy_topics — ntfy.sh topics
• pagerduty_integration_keys — PagerDuty Events API v2 routing keys
• twilio — Twilio SMS. You must supply your own Twilio sending source (account_sid, auth_token, and a verified from_phone) plus the destination to_phones. Unlike the hosted SolScope service, there is no shared sender — SMS only works if you provide your own Twilio account.
• smtp_email — SMTP email (host, port, username, password, from_email, to_emails, use_tls)

Recovery notifications

When a watcher's condition clears (e.g. your node comes back to the required version, or stops being delinquent), every configured channel automatically receives a RESOLVED: ... message. PagerDuty goes further: the original incident is auto-resolved via a stable dedup key (<identity>:<watcher>), so you don't have to close it by hand. This is always on — no extra config.

Cooldowns still apply while an alert persists (so you aren't spammed), but a recovery notification is sent exactly once on the transition back to healthy, and the cooldown resets so a re-occurrence alerts you immediately.

Install

Install into a virtual environment. This is the recommended approach everywhere, and on modern Debian/Ubuntu (PEP 668 "externally-managed-environment") it is required — a system-wide pip install will be refused.

# Create a venv (one-time). Anywhere is fine; this keeps it with the config.
python3 -m venv ~/.solscope-validator-watcher/venv

# Activate it, then install
source ~/.solscope-validator-watcher/venv/bin/activate
pip install solscope-validator-watcher

On Ubuntu you may first need sudo apt install python3-venv (and python3-pip).

After activating the venv, the solscope-validator-watcher command is on your PATH. You don't need to keep the venv activated for cron — see Run (cron), which records the venv's Python automatically.

If you prefer not to manage a venv yourself, pipx does it for you:

pipx install solscope-validator-watcher

From source (for development)

python3 -m venv .venv && source .venv/bin/activate
pip install -e .

Configure (the TUI)

The TUI is the primary entrypoint. Just run the command with no arguments:

solscope-validator-watcher

By default everything lives under ~/.solscope-validator-watcher/ (config.json, the cooldown state file, and watcher.log), so no root access is required. Use --config <path> to point at a different file.

The dashboard is a live status grid: each row is a validator, each column is a watcher, and every cell shows that check's current state — green ✓ ok, red ✗ alarm, dim — off (disabled), or yellow ! err (the check itself failed). All checks run on startup so you get an immediate picture, and you can re-run them any time with r. A live log pane at the bottom tails the cron log (~/.solscope-validator-watcher/watcher.log), so you can watch each scheduled run stream in as it happens.

• Enter on a row opens the editor for that validator; a adds a new one.
• In the editor: set cluster (or a custom RPC URL), identity/vote keys, toggle each watcher and its cooldown, and fill in notification channels.
  • Ctrl+S save · Ctrl+T send test notifications · Ctrl+D delete · Esc cancel
  • Test RPC button verifies connectivity (and reports the detected client, e.g. Agave/Firedancer, and version) before you save.
• Saving a validator automatically installs/refreshes the one-minute cron job, so monitoring is active without an extra step. Press c to (re)install it manually — handy if you move the virtualenv or the cron update was skipped.
• Press r to refresh the grid, q to quit.

Prefer to edit by hand? Copy config.example.json and edit it.

Custom RPC endpoint

By default each validator uses the public endpoint for its cluster (https://api.<cluster>.solana.com). Set a per-validator rpc_url to use your own RPC — the cluster value is still used for the version checks. Leave it null/blank for the public endpoint.

Run (cron)

The watchers run via the non-interactive run-once command, which the TUI's "install cron" action wires up for you. To do it manually (from inside the activated venv):

solscope-validator-watcher run-once          # run all validators once (used by cron)
solscope-validator-watcher install-cron      # install the one-minute cron job

cron does not inherit your activated virtualenv, so install-cron bakes the absolute path of the current Python interpreter into the cron line. As long as you run install-cron (or press c in the TUI) from inside the venv where you installed the package, the cron job will use that same venv automatically — no activation needed at run time. The installed line looks like:

* * * * * /home/solana/.solscope-validator-watcher/venv/bin/python -m validator_watcher run-once --config "..." >> "..." 2>&1

install-cron accepts --config, --python-bin (override the interpreter), and --log-file.

Recommended: high-availability setup

A monitor is only useful if it's running when something breaks — so don't rely on a single host to watch itself. Because one config can hold multiple validators, the most resilient setup is:

1. Add both your mainnet and testnet validators to the config (in the TUI, press a twice).
2. Install the watcher + cron on both hosts, each running that same config.

That way every validator is observed by two independent hosts. If one host goes down (the exact moment you most want an alert), the other host is still checking it and will fire the notification. The per-watcher cooldowns keep the two hosts from double-alerting you into noise.

License

MIT — see LICENSE.