activator: websocket connection dropped error log every minute

[snormore]

The activator logs are showing the following output every minute:

Aug 21 11:57:10 aws-dn-activator1 doublezero-activator[1175796]: [2025-08-21T11:57:10Z INFO  solana_pubsub_client::pubsub_client] receive error: WsError(Protocol(ResetWithoutClosingHandshake))
Aug 21 11:57:10 aws-dn-activator1 doublezero-activator[1175796]: [2025-08-21T11:57:10Z INFO  solana_pubsub_client::pubsub_client] websocket - exited receive loop
Aug 21 11:57:10 aws-dn-activator1 doublezero-activator[1175796]: [2025-08-21T11:57:10Z WARN  solana_pubsub_client::pubsub_client] unable to unsubscribe from websocket

This seems to be just the RPC server dropping the connection when there is no activity for some period of time (1 minute in this case), so it's both misclassified as an error and very noisy, making it difficult to surface and alert on real errors from the activator.

The activator (or rust SDK client specifically), should be updated to not log these as errors, warnings or even info-level, and to just reconnect as part of normal operations of being subscribed to the websocket RPC endpoint, or send keepalives to avoid disconnection every minute but even in that case the connection will be dropped sometimes with the expectation of reconnecting if needed.

Related to #1058