Bubble-up: Increasing utilization in modern warehouse scale computers via sensible co-locations

By Jason Mars, Lingjia Tang, Robert Hundt, Kevin Skadron, Mary Lou Soffa. Micro, 2011

649 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Hardware and Architecture, Information Systems

View Paper: https://www.semanticscholar.org/paper/bd9d8d5fe2bdb6e17cf6bf23b5ca24838bbc64bf

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Hardware and Architecture

In Information Systems

Citations in the corpus

Large-scale cluster management at Google with Borg (2015, background): “Even so, occasional low-level resource interference (e.g., memory bandwidth or L3 cache pollution) still happens, as in [60, 83].”
SoftSKU: Optimizing Server Architectures for Microservice Diversity @Scale (2019, background): “Other works reduce coscheduled job interference [109–114] or schedule them in a machine characteristics-aware manner [115–118].”
Heracles: Improving resource efficiency at scale (2015, methodology): “Bubble-flux is an online scheme that detects memory pressure and finds colocations that avoid interference on latencysensitive workloads [79, 51].”
Seer: Leveraging Big Data to Navigate the Complexity of Performance Debugging in Cloud Microservices (2019, background): “A second line of work tries to identify resources that will allow a new, potentially-unknown application to meet its performance (throughput or tail latency) requirements [29, 31, 32, 34, 66, 68, 95].”
Towards energy proportionality for large-scale latency-critical workloads (2014, methodology): “Alternatively, we can temporarily delay tasks to create suf.ciently long periods of idleness so that deep sleep modes are effective [28, 29].”
Efficient virtual memory for big memory servers (2013, background): “Thus, they often run either exclusively or with a few non-interfering tasks to guarantee high performance [29], low latency and performance predictability.”
PARTIES: QoS-Aware Resource Partitioning for Multiple Interactive Services (2019, background): “First, there are cluster schedulers that infer the expected interference of a given application colocation [13, 18, 20–22, 24, 33, 45, 58, 59, 61, 62], and either adjust allocations at runtime, or completely disallow resource sharing when the predicted latency violates QoS.”
Planaria: Dynamic Architecture Fission for Spatial Multi-Tenant Acceleration of Deep Neural Networks (2020, background): “There is a large swath of related work on multi-tenancy for CPUs [36, 66, 97–103] and GPUs [29, 35, 37–42, 66, 104–110] due to its vitality for cloud-scale computing.”
Quasar: resource-efficient and QoS-aware cluster management (2014, background): “Naı̈vely co-locating these services with low-priority, batch tasks that consume any idling resources can lead to unacceptable latencies, even at low load [41].”