Taming Server Memory TCO with Multiple Software-Defined Compressed Tiers

TL;DR

TierScape introduces multiple software-defined compressed memory tiers to tame memory TCO in data centers, extending beyond the traditional $2$-Tier model. By combining diverse compression algorithms, allocators, and backing media, TierScape creates a rich tier space that enables aggressive savings for warm and cold data while preserving performance. It pairs two data-placement strategies—a waterfall model and an analytical ILP-based model with a tunable knob $\alpha$—to dynamically migrate data across tiers based on real-time access profiles, implemented via kernel changes and the TS-Daemon. Experimental results on real workloads show substantial memory-TCO savings (up to $64.10\%$) with competitive or improved performance and acceptable tail-latency impacts, suggesting practical value for SLA-aware data-center deployments.

Abstract

Memory accounts for 33 - 50% of the total cost of ownership (TCO) in modern data centers. We propose a novel solution to tame memory TCO through the novel creation and judicious management of multiple software-defined compressed memory tiers. As opposed to the state-of-the-art solutions that employ a 2-Tier solution, a single compressed tier along with DRAM, we define multiple compressed tiers implemented through a combination of different compression algorithms, memory allocators for compressed objects, and backing media to store compressed objects. These compressed memory tiers represent distinct points in the access latency, data compressibility, and unit memory usage cost spectrum, allowing rich and flexible trade-offs between memory TCO savings and application performance impact. A key advantage with ntier is that it enables aggressive memory TCO saving opportunities by placing warm data in low latency compressed tiers with a reasonable performance impact while simultaneously placing cold data in the best memory TCO saving tiers. We believe our work represents an important server system configuration and optimization capability to achieve the best SLA-aware performance per dollar for applications hosted in production data center environments. We present a comprehensive and rigorous analytical cost model for performance and TCO trade-off based on continuous monitoring of the application's data access profile. Guided by this model, our placement model takes informed actions to dynamically manage the placement and migration of application data across multiple software-defined compressed tiers. On real-world benchmarks, our solution increases memory TCO savings by 22% - 40% percentage points while maintaining performance parity or improves performance by 2% - 10% percentage points while maintaining memory TCO parity compared to state-of-the-art 2-Tier solutions.

Figures (13)

• Figure 1: Memcached on a $2$-Tier system (DRAM + a single compressed tier): conservatively placing 20% cold data in the compressed tier limits the memory TCO savings to 11% with a 9.5% slowdown. Placing around 50% of data (including cold and some warm data) in the compressed tier results in 16% memory TCO savings and 13.5% slowdown. An aggressive approach that places around 80% of data (including cold and most of the warm data) in the compressed tier results in 32% memory TCO savings and 20% slowdown.
• Figure 2: Data placement options in $2$-Tier and N-Tier systems
• Figure 3: Characterization results for 12 different software-defined compressed tiers for dicken and nci data sets. Encoding: ZS, ZB refers to zsmalloc and zbud pool managers, respectively. L4, LO, DE refers to lz4, lzo, and deflate compression algorithms, respectively. DR, OP: refers to DRAM and Optane optane as the backing storage media, respectively.
• Figure 4: Page placement with the N-Tier waterfall model
• Figure 5: Page placement with the N-Tier analytical model.