Partition and Isolate: Approaches for Consolidating HPC and Commodity Workloads Jack Lange Assistant Professor University of Pittsburgh

Summary Commodity and HPC systems have been converging – Commodity off the shelf components – Linux based HPC systems – Cloud computing Problem: Real HPC applications need HPC environments – Tightly coupled, massively parallel, and synchronized – Current services must provide dedicated HPC systems Can we co-host HPC applications on commodity systems? Dual Stack Approach – Provision the underlying software stack along with application – Commodity stack should handle commodity applications – HPC stack can provide HPC environment

Current systems do support this, but… Interference still exists inside the system software – Inherent feature of commodity systems Cores Socket 1 Memory Cores Socket Memory HPC Partition Commodity Partition User Space Partitioning

HPC vs. Commodity Systems Commodity systems have fundamentally different focus than HPC systems – Amdahl’s vs. Gustafson’s laws – Commodity: Optimized for common case HPC: Common case is not good enough – At large (tightly coupled) scales, percentiles lose meaning – Collective operations must wait for slowest node – 1% of nodes can make 99% suffer – HPC systems must optimize outliers (worst case)

5 High Performance Computing (HPC) Large scale simulations to solve Big Problems

Dual Stack Approach Partition – Segment the underlying hardware resources – Assign them to exclusively to specific workloads Isolate – Prevent interference from other workloads – Hardware: partitions must be course grained – Software: eliminate shared state Implementation – Independent system software running on isolated resources

HPC in the cloud Clouds are starting to look like supercomputers… – Are we seeing a convergence? Not yet – Noise issues – Poor isolation – Resource contention – Lack of control over topology Very bad for tightly coupled parallel apps – Require specialized environments that solve these problems Approaching convergence – Vision: Dynamically partition cloud resources into HPC and commodity zones – This talk: partitioning compute nodes with performance isolation

Commodity VMMs Virtualization is considered an “enterprise” technology – Designed for commodity environments – Fundamentally different, but not wrong! Example: KVM architecture issues – Userspace handlers – Fairly complex memory management – Locking and periodic optimizations – Presence of system noise

9 Palacios VMM OS-independent embeddable virtual machine monitor – Established compatibility with Linux, Kitten, and Minix Specifically targets HPC applications and environments – Consistent performance with very low variance Deployable on supercomputers, clusters (Infiniband/Ethernet), and servers – 0-3% overhead at large scales (thousands of nodes) VEE 2011, IPDPS 2010, ROSS Open source and freely available

Palacios/Linux Palacios/Linux provides lightweight and high performance virtualized environments – Internally manages dedicated resources Memory and CPU scheduling – Does not bother with “enterprise features” Page sharing/merging, swapping, overcommitting resources Palacios enables scalable HPC performance on commodity platforms

VMM Comparison Primary difference: Consistency – Requirement for tightly coupled performance at large scale Example: KVM nested paging architecture – Maintains free page caches to optimize performance Requires cache management – Shares page tables to optimize memory usage Requires synchronization VMM% of exitsMeanStd Dev# NPFS KVM52% ,265,156 Palacios50% ,872,017

Hardware Palacios VMM HPC Linux Linux Kernel HPC Application Palacios Resource Managers KVM Linux Module Interface Commodity Linux Commodity Application(s) Dual Stack Architecture Partitioning at the OS level Enable cloud to host both commodity and HPC apps – Each zone optimized for the target applications

Evaluation Goal: Measure VM isolation properties Partitioned a single node into HPC and commodity zones – Commodity Zone: Parallel Kernel compilation – HPC Zone: Set of standard HPC benchmarks – System: Dual 6-core AMD Opteron with NUMA topology Linux guest environments (HPC and commodity) Important: Local node only – Does not promise good performance at scale – But, poor performance will magnify at large scales

Results Palacios delivers consistent performance Commodity VMMs degrade with contention MiniFE: Unstructured implicit finite element solver Mantevo Project --

Discussion A dual stack approach can provide HPC environments on commodity clouds – HPC and commodity workloads can dynamically share resources – HPC requirements can be met without fully dedicated resources Networking is still an open issue – Need mechanisms for isolation and partitioning – Need high performance networking architectures 1Gbit is not good enough 10Gbit is good, Infiniband is better – Need control over placement and topologies

Multi-stack Operating Systems Future Exascale Systems are moving towards in situ organization Applications traditionally have utilized their own platforms Visualization, storage, analysis, etc Everything must now collapse onto a single platform

What this means for the OS At Petascale we could optimize each environment separately – Each had their own OS and hardware At Exascale workloads will be co-located – Can a single OS handle all workloads effectively? Probably Not – Each has different resource requirements and behaviors – Exascale will need to support multiple OS environments on the same hardware

Current Supercomputer OS architectures 18 Source: e:Operating_systems_used_on _top_500_supercomputers.svg LINUX UNIX

19 Will Linux continue to dominate? An open question at this point – Exascale systems may mark a radical departure from traditional architectures Kitten: Open-source Lightweight Kernel from Sandia – Minimal compute node OS Provides mostly Linux-compatible user environment – Supports unmodified compiler toolchains and ELF executables But doesn’t include the enterprise Linux features – Simple memory management, application managed I/O, etc

Hardware Palacios VMM Kitten Linux Kernel HPC Application Palacios Resource Managers Linux Linux Module Interface Commodity Application(s) Dual Stack Architecture Provide LWK environment on a commodity system – Each zone optimized for the target applications

Palacios/Kitten provides higher memory throughput than Linux – 400 MB/s (4.74%) Palacios/Kitten provides more consistent memory performance than Linux – 340 MB/s lower standard deviation Stream 21

Selfish Detour Virtualized Kitten 22 LinuxVirtual Kitten Palacios/Kitten can reduce noise from Linux Eliminates Periodic timer interrupts

Better than native? Results are preliminary – Followed best practices for configuring Linux – Didn’t try to optimize VMM performance Virtualization can improve system performance – when the system is running a commodity OS B. Kocoloski, J. Lange, Better than Native: Using Virtualization to Improve Compute Node Performance, (ROSS 2012)

Beyond Virtualization Virtualization imposes overhead – Power: requires transistors – Performance: small, but present – Interference: Still some dependencies on host OS Might not be available on exascale hardware Can we provide native partitioning? – We think so – Linux provides the ability to dynamically remove resources (CPUs, memory, etc) – These can be taken over by a second OS

Hardware Palacio VMM Kitten HPC Application Linux Commodity Application(s) Dual Stack Architecture Provide LWK environment on a commodity system – Each zone optimized for the target applications

Approach OS partition created via offlined resources – CPUs, memory, PCI devices Secondary OS “booted” on offline resources Issues: – OS initialization Boot process Resource discovery – Coordination and communication – Security and safety

Dual Stack Memory Maybe we don’t need to provide an entirely separate OS – Instead selectively manage some resources Dual stack memory – Provide a separate memory management layer to Linux Features – Selectively manage heap per application – Provide applications with direct control over memory layout – Transparently back memory using large pages – Without overhead added by Linux

Hardware HPC Application Linux Commodity Application(s) Dual Stack Architecture Provide LWK memory manager on a commodity OS Memory Management

Conclusion Commodity systems are not designed to support HPC workloads – Different requirements and behaviors than commodity applications A multi stack approach can provide HPC environments in commodity systems – HPC requirements can be met without separate physical systems – HPC and commodity workloads can dynamically share resources – Isolated system software environments

Thank you Jack Lange Assistant Professor University of Pittsburgh