Network operators increasingly rely on in-network functionality to make their networks manageable and economically viable. These middleboxes make the end-to-end path for traffic more opaque by making assumptions about the traffic passing through them. This has led to an ossification of the Internet protocol stack: new protocols and extensions can be difficult to deploy when middleboxes do not understand them [Honda11]. PATHspider is a software measurement tool for active measurement of Internet path transparency to transport protocols and transport protocol extensions, that can generate raw data at scale to determine the size and shape of this problem.
The A/B testing measurement methodology used by PATHspider is simple: We perform connections from a set of observation points to a set of measurement targets using two configurations. A baseline configuration (A), usually a TCP connection using kernel default and no extensions, tests basic connectivity. These connections are compared to the experimental configuration (B), which uses a different transport protocol or set of TCP extensions. These connections are made as simultaneously as possible, to reduce the impact of transient network changes.
Since PATHspider 2.0, it is also possible to perform more than one B test optionally performing an A test between each B test in order to revalidate the path. This can also be used to “prime” the path for a follow up connection if it is desirable to have devices on the path hold state before performing the test.
PATHspider is a generalized version of the ecnspider tool, used in previous studies to probe the paths from multiple vantage points to web-servers [Trammell15] and to peer-to-peer clients [Gubser15] for failures negotiating Explicit Congestion Notification (ECN) [RFC3168] in TCP.
As a generalized tool for controlled experimental A/B testing of path impairment, PATHspider fills a gap in the existing Internet active measurement software ecosystem. Existing active measurement platforms, such as RIPE Atlas [RIPEAtlas], OONI [Filasto12], or Netalyzr [Kreibich10], measure absolute performance and connectivity between a pair of endpoints under certain conditions. The results obtainable from each of these can be compared to each other to simulate A/B testing. However, the measurement data from these platforms provide a less controlled view than can be achieved with PATHspider, given coarser scheduling of measurements in each state.
Given PATHspider’s modular design and implementation in Python, plugins to perform measurements for any transport protocol or extension are easy to build and can take advantage of the rich Python library ecosystem, including high-level application libraries, low-level socket interfaces, and packet forging tools such as Scapy.
The PATHspider architecture has four components, illustrated in
the diagram below the
observer and the
combiner. Each component is implemented as one or more
threads, launched when PATHspider starts.
For each target hostname and/or address, with port numbers where appropriate, PATHspider enqueues a job, to be distributed amongst the worker threads when available. Each worker performs one connection with the “A” configuration and one connection with the “B” configuration. The “A” configuration will always be connected first and serves as the base line measurement, followed by the “B” configuration. This allows detection of hosts that do not respond rather than failing as a result of using a particular transport protocol or extension. These sockets remain open for a post-connection operation.
Some transport options require a system-wide parameter change, for example enabling ECN in the Linux kernel. This requires locking and synchronisation. Using semaphores, the configurator waits for each worker to complete an operation and then changes the state to perform the next batch of operations. This process cycles continually until no more jobs remain.
In a typical experiment, multiple workers (on the order of hundreds) are active, since much of the time in a connection test is spent waiting for an answer from the target or a timeout to fire. Where it is possible to peform the tests without a system-wide configuration it is possible to disable the semaphores to increase the speed of the test.
In addition, packets are separately captured for analysis by the observer using Python bindings for libtrace. First, the observer assigns each incoming packet to a flow based on the source and destination addresses, as well as the TCP, UDP or SCTP ports when available. The packet and its associated flow are then passed to a function chain. The functions in this chain may be simple functions, such as counting the number of packets or octets seen for a flow, or more complex functions, such as recording the state of flags within packets and analysis based on previously observed packets in the flow. For example, a function may record both an ECN negotiation attempt and whether the host successfully negotiated use of ECN.
Path conditions are generated for the path to each target to determine whether or not connectivity breakage has occured, or other conditions that may lead to more subtle breakage.
PATHspider plugins are built by extending an abstract class that
implements the core behaviour, with functions for the
configurator, workers, observer, and merger. There are three main abstract
classes that can be extended by plugins:
Depending on the type of plugin being created, these abstract classes are extended to include logic for generating the active measurement traffic.
Plugins can implement arbitrary functions for the observer function chain, or reuse library functions for some functionality. These track the state of flows and build flow records for different packet classes: The first chain handles setup on the first packet of a new flow. Separate chains for IP, TCP and UDP packets allow different behaviours based on the IP version and transport protocol.
The final plugin function is the combiner function. This takes a list of merged job records and flow records to produce “path conditions” before passing the final job record back to PATHspider for output.