In-memory database management systems (DBMSs) outperform disk-oriented systems for on-line transaction processing (OLTP) workloads. But this improved performance is only achievable when the database is smaller than the amount of physical memory available in the system. To overcome this limitation, some in-memory DBMSs can move cold data out of volatile DRAM to secondary storage. Such data appears as if it resides in memory with the rest of the database even though it does not. Although there have been several implementations proposed for this type of cold data storage, there has not been a thorough evaluation of the design decisions in implementing this technique, such as policies for when to evict tuples and how to bring them back when they are needed. These choices are further complicated by the varying performance characteristics of different storage devices, including future non-volatile memory technologies. We explore these issues in this paper and discuss several approaches to solve them. We implemented all of these approaches in an in-memory DBMS and evaluated them using five different storage technologies. Our results show that choosing the best strategy based on the hardware improves throughput by 92-340% over a generic configuration.
Function as a Service (FaaS) and the associated serverless computing paradigm alleviates users from resource management and allows cloud platforms to optimize system infrastructure under the hood. Despite significant advances, FaaS infrastructure still leaves much room to improve performance and resource efficiency. We argue that both higher performance and resource efficiency are possible --- while maintaining secure isolation --- if we are willing to revisit the FaaS programming model and system software design. We propose Dandelion, a clean-slate FaaS system that rethinks the programming model by treating serverless functions as pure functions, thereby explicitly separating computation and I/O. This new programming model enables a lightweight yet secure function execution system. It also makes functions more amenable to hardware acceleration and enables dataflow-aware function orchestration. Our initial prototype of Dandelion achieves 45× lower tail latency for cold starts compared to Firecracker. For 95% hot function invocations, Dandelion achieves 5× higher peak throughput.
This paper introduces a software policy for memory management in heterogeneous memory systems in order to improve the trade-offs between performance and power consumption, while attempting to make the best use of different characteristics of the underlying memory technologies. In this policy, the operating system and the application co-schedule page management in order to make informed decisions about page allocation and migration. Software-Controlled 2- Level Memory (Soft2LM) is a hardware-agnostic approach for efficient usage of heterogeneous memory that allows for region-based allocations, migrations, and application advice. We include analysis of the access characteristics of the PARSEC 3.0 suite of benchmarks, both as motivation for software-guided intelligent page placement as well as for understanding where application advice can make a difference. Our evaluation running PARSEC on our Soft2LM Linux kernel shows an average 4.7x improvement over comparable RAMDISK- based swapping. Even when migrating 100 and 1000 pages per second between regions, Soft2LM performs comparably to a stock system with twice the available DRAM, showing a 1.8\% and 0.7\% improvement, respectively.
With trends in mixed-signal systems-on-chip indicating increasingly extreme scaling of device dimensions and higher levels of integration, the tasks of both design and device validation is becoming increasingly complex. Post-silicon validation of mixed-signal/RF systems provides assurances of functionality of complex systems that cannot be asserted by even some of the most advanced simulators. We introduce RAVAGE (from “random;” “validation;” and “generation”), an algorithm for generating stimuli for post-silicon validation of mixed-signal systems. The approach of RAVAGE is new in that no assumption is made about any design anomaly present in the DDT; but rather, the stimulus is generated using the DUT itself with the objective of maximizing the effects of any behavioral differences between the DUT (hardware) and its behavioral model (software) as can be seen in the differences of their response to the same stimulus. Stochastic test generation is used since the exact nature of any behavioral anomaly in the DUT cannot be known a priori. Once a difference is observed, the model parameters are tuned using nonlinear optimization algorithms to remove the difference between its and the DUT's responses and the process (test generation→tuning) is repeated. If a residual error remains at the end of this process that is larger than a predetermined threshold, then it is concluded that the DUT contains unknown and possibly malicious behaviors that need further investigation. Experimental results on an RF system (hardware) are presented to prove feasibility of the proposed technique.
There are numerous ways to control power usage of modern systems. Dynamic voltage and frequency scaling (DVFS) and C- and P-states are commonly available in hardware, and operating systems have fine-grained control of these states. Unfortunately, most DVFS systems simply consider CPU utilization when determining which P-state to enter. Running average power limit (RAPL), introduced by Intel in their Sandy Bridge line of processors, allows researchers and system designer to obtain detailed estimates of energy consumption by the core, uncore and DRAM. In this paper we use Linux's perf_events subsystem to measure a number of commonly cited performance metrics while running the SPEC CPU2006 benchmarks in addition to calculating average power using RAPL registers. This paper shows initial promising results correlating DRAM memory consumption to stall-cycle ratio.
With the computational systems of even embedded devices becoming ever more powerful, there is a need for more effective and pro-active methods of dynamic power management. The work presented in this paper demonstrates the effectiveness of a reinforcement-learning based dynamic power manager placed in a software framework. This combination of Q-learning for determining policy and the software abstractions provide many of the benefits of co-design, namely, good performance, responsiveness and application guidance, with the flexibility of easily changing policies or platforms. The Q-learning based Quality of Service Manager (2QoSM) is implemented on an autonomous robot built on a complex, powerful embedded single-board computer (SBC) and a high-resolution path-planning algorithm. We find that the 2QoSM reduces power consumption up to 42% compared to the Linux on-demand governor and 10.2% over a state-of-the-art situation aware governor. Moreover, the performance as measured by path error is improved by up to 6.1%, all while saving power.