Open jterzis opened 3 months ago
There is #619 to add provider status to /status for monitoring, current take on supporting fallback is that if this functionality is required RPC proxy should be run, so putting it on the backburner. If there are problems with running proxies we'll reevaluate bringing this functionality into the node.
Is your feature request related to a problem? Please describe. Earlier this week, the network experienced an issue due to a faulty base network url. Infura's origin was experiencing downtime and as a result, several stream nodes in the mainnet network were in effect down for reads/writes. The operators solved the issue by restarting the stream node software under a different network url provider.
Describe the solution you'd like A simple fallback URL that operators can inject into the runtime such that if the BASECHAIN__NETWORKURL experiencing downtime over a specific time period, the node seamlessly switches connectivity to the fallback.
Describe alternatives you've considered Alternatives include having operators run an RPC proxy against multiple upstreams that checks health and actively diverts traffic to the most alive upstream.