This page describes how Paragon responds to incidents on cloud and managed on-premise installations. On unmanaged on-premise, you own monitoring, incident response, and patching; Paragon can advise if you share diagnostics.For managed installations, hundreds of metrics are collected with alarms for Paragon’s support team. You can learn more here.
System Metrics and Alarms
We proactively respond to deviations and often remediate misconfigurations, errors, and costs before customers or their users are made aware.If you detect issues with your environment, please contact our team, and we’ll prioritize fixing and patching your environment as soon as possible.
In almost every case, we’re notified within seconds when a managed installation lacks resources. Autoscaling is typically enabled within predefined limits. If a resize would materially affect monthly costs, we notify your team and request approval for the installer to update the system.
When high error rates are detected, we’ll immediately investigate the source of error rates and take the appropriate action. In many cases, it’s a new workflow that’s being tested or misconfigured, and no action is needed.Generally, releases are tested in our cloud environment and confirmed to be stable before deploying to enterprise environments, so rollbacks due to high error rates in on-premise environments rarely happen.
If we detect an exploit, your team is notified immediately and a Paragon response team is formed. If you become aware of an exploit, notify us immediately.
Every microservice is configured with horizontal pod scalers, services, and health checks to enforce uptime. If services are regularly going offline due to high load or a lack of resources, please report it to the support team, and we’ll make sure to investigate and optimize your installation.
If you find a bug or regression within your installation, please report it to the support team, and we’ll prioritize patching and updating your system as soon as possible.
Was this page helpful?
Assistant
Responses are generated using AI and may contain mistakes.