Uptime monitoring dashboard for the production GrapheneOS services is available at nodeping.com/reports/status…. Alerts are also posted publicly in our #infra:grapheneos.org Matrix room when a check is added, starts failing or stops failing. It's often not really an outage.
For example, today, the TLS checks for attestation.app detected the certificate would expire in less than 15 days. This ended up being because the removal of a legacy subdomain several weeks ago broke automatic certbot removal. It can do much more advanced checks though.
For example, our "matrix.grapheneos.org HTTPS Federation API" check fetches matrix.grapheneos.org/_matrix/key/v2… and makes sure that it contains "server_name"\s*:\s*"grapheneos.org". We also have one for the Matrix client API. Checks also also be set to roam around internationally.
At the moment, we either set the checks to use a specific location for more useful latency monitoring or we set them to roam worldwide. Here's an example of a worldwide check:
nodeping.com/reports/status…
It's also a supplemental way of testing GeoDNS for the services we use it.
nodeping.com/reports/status…
It's also a supplemental way of testing GeoDNS for the services we use it.
Since we primarily use OVH for our services, we make sure we're using non-OVH-based check locations for the non-roam checks. This is also part of why using an external service makes more sense than self-hosting and we don't send alerts to our own mail server for the same reason.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
