WS connection fails after few received "newHeads" messages

[DimitarTAtanasov]

Hi, I am running blockscout locally using Docker-compose Deployment.
Few block numbers are being received via the websocket connection as expected and processed by apps/indexer/lib/indexer/block/realtime/fetcher.ex -> def handle_info( {subscription, {:ok, %{"number" => quantity}}},, the additional block information is fetched by start_fetch_and_import and he blocks are being inserted.
After few blocks I am hitting the fallback function

         def handle_info(_, state) do
         {:noreply, state}
          end

After hitting the fallback function, the new blocks are being fetched via the fallback polling mechanism.

I am trying to figure out, why the websocket connections fails.

NOTES:

• I don't think the websocket connection itself disconnects due to network issues as this happens even if I am connected to locally hosted geth node.
• It doesn't look I am receiving any malformed or unexpected messages through the WS(I am adding logs in the websocket client handle_frame and handle_responce funcs and the messages appear to be correct and processed)
• Using the following env vars

        ETHEREUM_JSONRPC_HTTP_URL: http://host.docker.internal:8545/
        ETHEREUM_JSONRPC_TRACE_URL: http://host.docker.internal:8545/
        ETHEREUM_JSONRPC_WS_URL: ws://host.docker.internal:8546/
        INDEXER_CATCHUP_BLOCK_INTERVAL: '50ms'
        ETHEREUM_JSONRPC_FALLBACK_WS_URL: ws://host.docker.internal:8546/

Any help resolving this issue will be appreciated.

[ekawibisono]

Hello, please you can ask questions on discord community blockscout https://discord.gg/blockscout