Analysing Mechanisms for Virtual Channel Management in Low-Diameter networks

TL;DR

This paper investigates buffer management policies in low-diameter networks like HyperX and Dragonfly+, revealing their critical impact on performance and deadlock avoidance, and proposes solutions to improve network reliability.

Contribution

It characterizes the impact of buffer management policies on performance and deadlock prevention in low-diameter networks and offers practical solutions.

Findings

Incorrect buffer policies can cause up to 90% performance drop.

Proper buffer management is essential for deadlock-free routing.

The study proposes effective buffer management solutions.

Abstract

To interconnect their growing number of servers, current supercomputers and data centers are starting to adopt low-diameter networks, such as HyperX, Dragonfly and Dragonfly+. These emergent topologies require balancing the load over their links and finding suitable non-minimal routing mechanisms for them becomes particularly challenging. The Valiant load balancing scheme is a very popular choice for non-minimal routing. Evolved adaptive routing mechanisms implemented in real systems are based on this Valiant scheme. All these low-diameter networks are deadlock-prone when non-minimal routing is employed. Routing deadlocks occur when packets cannot progress due to cyclic dependencies. Therefore, developing efficient deadlock-free packet routing mechanisms is critical for the progress of these emergent networks. The routing function includes the routing algorithm for path selection and the buffers management policy that dictates how packets allocate the buffers of the switches on their paths. For the same routing algorithm, a different buffer management mechanism can lead to a very different performance. Moreover, certain mechanisms considered efficient for avoiding deadlocks, may still suffer from hard to pinpoint instabilities that make erratic the network response. This paper focuses on exploring the impact of these buffers management policies on the performance of current interconnection networks, showing a 90\% of performance drop if an incorrect buffers management policy is used. Moreover, this study not only characterizes some of these undesirable scenarios but also proposes practicable solutions.