Columbo: Low Level End-to-End System Traces through Modular Full-System Simulation

TL;DR

This paper introduces Columbo, a system that enables detailed, end-to-end tracing of cloud systems at the hardware and software level using full-system simulation, overcoming physical testbed limitations.

Contribution

It presents a novel approach to system tracing by leveraging full-system simulation to obtain fine-grained, in-depth insights into cloud system behavior.

Findings

Enables detailed system traces without affecting real systems.

Provides insights into sub-microsecond interactions between hardware and software.

Facilitates analysis of complex cloud system performance.

Abstract

Fully understanding performance is a growing challenge when building next-generation cloud systems. Often these systems build on next-generation hardware, and evaluation in realistic physical testbeds is out of reach. Even when physical testbeds are available, visibility into essential system aspects is a challenge in modern systems where system performance depends on often sub-$\mu s$ interactions between HW and SW components. Existing tools such as performance counters, logging, and distributed tracing provide aggregate or sampled information, but remain insufficient for understanding individual requests in-depth. In this paper, we explore a fundamentally different approach to enable in-depth understanding of cloud system behavior at the software and hardware level, with (almost) arbitrarily fine-grained visibility. Our proposal is to run cloud systems in detailed full-system simulations, configure the simulators to collect detailed events without affecting the system, and finally assemble these events into end-to-end system traces that can be analyzed by existing distributed tracing tools.