After receiving an error reading from an sqs queue, propono spins and never recovers. about propono

[malcyL]

The code where this is happening is here:

    def listen
      raise ProponoError.new("topic_id is nil") unless @topic_id
      loop do
        unless read_messages
          sleep 10
        end
      end
    end

    private

    def read_messages
      response = sqs.receive_message( queue_url, {'MaxNumberOfMessages' => 10} )
      messages = response.body['Message']
      if messages.empty?
        false
      else
        messages.each { |msg| process_sqs_message(msg) }
      end
    rescue
      Propono.config.logger.error "Unexpected error reading from queue #{queue_url}"
      Propono.config.logger.error $!
    end

We are rescuing the error in read_messages and then continuing to loop in listen, after a delay of 10 seconds. This continues to spin until the process is stopped, but on restart we are able to read from the sqs queue without any problem.

[iHiD]

@malcyL Is this still an issue or did we fix it?

[malcyL]

@iHiD This has never been reproduced, in live or test.

It's possible that whatever caused this has been fixed - not a great answer I know.

We left the issue open so we had a record of the failure. It's now 3 months later. We have some significant usage of this in live now. With no re occurrences, and still being unable to reproduce it. I propose closing it.