config.yml

### Instattack Configuration Settings
### TODO: Turn Into Python File

instattack:
    debug: true

    log.logging:
        level: 'debug'
        log_request_errors: true
        log_proxy_queue: true

connection:
    limit_per_host: 0
    connection_timeout: 14
    connection_limit: 0


login:

    passwords:
        # The number of concurrent passwords to try at the same time.
        passwords_batch_size: 10

        generator:
            # TODO: Add validation to make sure all elements of nested lists are
            # integers.
            capitalize_at_indices: [0, 1, [0, 1]]

            alterations:
                before: false
                after: true

            numerics:
                before: false
                after: true
                birthday:
                    provided: true
                    all: false
                    # TODO: Make required based on the value of `all`.
                    start_year: 1991
                    end_year: 2000

    attempts:
        # The number of concurrent requests to make for each password.
        attempts_batch_size: 20
        save_method: 'end'


proxies:
    save_method: 'end'
    countries: ['US']

    train:
        batch_size: 50
        # timeout: 10  # Replaces Timeout for Connection

    # Settings Here Only Matter if Collect = True
    # Do we want to add a limit to the broker collection?
    broker:
        # The maximum number of concurrent checks of proxies
        broker_max_conn: 200

        # The maximum number of attempts to check a proxy
        # Note that the lower the value of max_tries, the faster the broker will return
        # proxies, but they will not contain as much information to gauge their
        # relative reliability and speed.
        broker_max_tries: 2

        # Maximum amount of time to wait for broker to return a proxy before failing.
        broker_timeout: 5

    pool:
        # The maximum amount of time to wait for a proxy from the queue before failing.
        pool_timeout: 10
        collect: false

        # 3 Purposes:
        #   (1) Initial Population of Confirmed Queue
        #   (2) Whether or Not Evaluation Can be Ignored
        #   (3) Whether or not Error Causes Confirmed Proxy to be Removed

        # Right now this checks if both are satisfied, but we might want the ability
        # to check if either is specified, and over what range.
        confirmation:
            threshold: 2
            horizon: 5
            threshold_in_horizon: 1

        # We do not need the confirmed fields since they already are factored
        # in based on the separate queues.
        priority: [
            [-1, ['requests', 'active', 'success']],
            [1, ['error_rate', 'active']],
            [-1, ['requests', 'historical', 'success']],
            [1, ['error_rate', 'historical']],
            [1, ['avg_resp_time']]
        ]

        # The amount of time to wait before using a proxy that raises an error
        # that does not indicate an invalid proxy.
        # Each additional increment is calculated as increment = start*e^(c*x), where
        # x is the number of times the request has failed consecutively.
        # After the max value has been reached via incrementation, the
        # proxy will be discarded.
        timeouts:
            too_many_requests:
                increment: 5
                start: 5
                max: 40

            too_many_open_connections:
                increment: 5
                start: 5
                max: 40

        limits:
            # The maximum response time in seconds. If proxy.avg_resp_time exceeds this
            # value, proxy will be removed from the pool.
            # TODO: Allow Historical & Current
            resp_time: null

            # The maxmium number of allowed requests for each proxy per attack, before
            # being removed from pool.  For LARGER attacks, this should probably be
            # negated, since we do not want to discard good proxies.
            requests:
                all:
                    historical: null
                    active: null
                success:
                    historical: null
                    active: null
                fail:
                    historical: null
                    active: null

            # The maximum overall error rate for a proxy to be put into or used from
            # the pool.
            # TODO: Start Measuring Historical & Active
            error_rate:
                historical: null
                active: null
                # The minimum number of requests required before error rate is non-zero.
                horizon: 5

            # The maximum number of allowed errors for a proxy before being
            # removed from the pool or not included in pool to begin with.
            errors:
                # Right now, too_many_requests counts as a general error, so
                # we probably do not want to limit the number of total errors.
                # We need to treat too_many_requests as a warning.
                # all:
                #     historical: 2
                #     active: 1
                connection:
                    historical: 2
                    active: 1
                response:
                    historical: 2
                    active: 1
                instagram:
                    historical: 1
                    active: 1
                ssl:
                    historical: 3
                    active: 2
                timeout:
                    # There seem to be a lot of proxies that do get confirmed results
                    # but also have a lot of historical timeout errors.
                    historical: 4
                    active: 2