Troubleshooting blackbox SDN control software with minimal causal sequences

Abstract

Software bugs are inevitable in software-defined networking control software, and troubleshooting is a tedious, time-consuming task. In this thesis we discuss how to improve control software troubleshooting by presenting a technique for automatically identifying a minimal sequence of inputs responsible for triggering a given bug, without making assumptions about the language or instrumentation of the software under test. We apply our technique to five open source SDN control platforms-Floodlight, NOX, POX, Pyretic, ONOS-and illustrate how the minimal causal sequences our system found aided the troubleshooting process. AcknowledgmentsMany thanks to the STS team for making this thesis possible:

A Precise and Expressive Lattice-theoretical Framework for Efficient Network Verification

“…RELATED WORK Similar to APV [27] and ddNF [28], #PEC has many potential applications in the field of network correctness. The literature in this field is vast and includes BGP configuration checking (e.g., [7], [44]- [51]), ACL misconfiguration detection (e.g., [52], [53]), firewall checking (e.g., [17], [18], [21], [54]), SDN verification (e.g., [20], [23], [55], [56]), testing (e.g., [2], [57]- [60]), debugging (e.g., [61], [62]), differential analysis (e.g., [63]), concurrency analysis (e.g [64], [65]), automatic repair (e.g., [66]- [68]), synthesis (e.g. [69]- [71]), programming languages (e.g.…”

Toward Consistent SDNs: A Case for Network State Fuzzing

“…Data Plane Logical Rules (R logical ) [26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] (P == R logical ) (P == P logical ) [5,[21][22][23][24][25] (P == P physical ) [20] (P logical == P physical ) control, and data plane -with their components. The management plane establishes the network-wide policy P, which corresponds to the network operator's intent.…”

“…Control plane solutions focus on verifying network-wide invariants such as reachability, forwarding loops, slicing, and black hole detection against high-level network policies both for stateless and stateful policies. This includes tools [26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] that monitor and verify some or all of the network-wide invariants by comparing the highlevel network policy with the logical rule set that translates to the logical path set at the control plane, i.e., P ≡ R logical or P ≡ P logical . These systems only model the network behavior which is insufficient to capture firmware and hardware bugs as "modelling" and verifying the control-data plane consistency are significantly different techniques.…”

“…In addition to the related work covered in §1 that includes the existing literature based on [17] and Table 1, we now will navigate the landscape of related works and compare them to PAZZ in terms of the Type-p and Type-a faults which cause inconsistency ( §2). The related work in the area of control plane [26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] either check the controller-applications or the control-plane compliance with the high-level network policy. These approaches are insufficient to check the physical data plane compliance with the control plane.…”

FloodGuard: A DoS Attack Prevention Extension in Software-Defined Networks

“…Some existing studies such as NICE [11], VeriCon [7] and [16] propose several methods to verify different features of control applications. In [26], the authors improve the performance of control software troubleshooting by using a minimal causal sequence of triggering events.…”