Modernising HPC Cluster Provisioning

Total Page:16

File Type:pdf, Size:1020Kb

Modernising HPC Cluster Provisioning Modernising HPC cluster provisioning Jordi Blasco (HPCNow!) Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Agenda 1 Motivation 2 Deployment Provisioning 3 Image Provisioning State-full NFSROOT State-less Semi State-less / State-lite 4 Improvements in diskless provisioning BeeGFS-root Lustre-root 5 Introducing OverlayFS root SquashFS Cluster File System Client 6 Conclusions Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Motivation Main Objectives • Flexibility • Resilience • Scalability • Online changes • Fast provisioning • DevOps & CI friendly Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Deployment Based Provisioning Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Deployment Based Provisioning Standard Pipeline • The system boots from PXE • Unattended installation based on OS dependent template (Kickstart|Autoyast|Preseed) • The OS is installed on local hard disks • Reboot the system in order to boot from local disk • (optional) Configuration manager (Puppet, cfengine, ansible) completes the setup and ensures consistency Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Deployment Based Provisioning Pros • Flexible management • DevOps & CI friendly Cons • Local disks are not reliable • Long time required to deploy a reasonably large cluster • Not easy to apply changes online • Limited scalability • Risk of inconsistency if configuration manager is not used Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Image Based Provisioning Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Local Disk (statefull) Standard Pipeline • The image is generated on a "golden" node • The image is uploaded to an NFS server. • (optional) generate a torrent file to propagate the image faster. • Boot the system from PXE • Load a rigid and limited OS with the cloning software (i.e. SystemImager) • Clone the local disk with the OS image. • Reboot the system from local disk. Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Local Disk (statefull) Pros • Easy to manage • Consistent configuration across the cluster • Still room for configuration managers Cons • Local disk required but not reliable or expensive • Not easy to apply changes into the disk image • Not easy to apply changes online • Takes reasonable amount of time to update the image and test it. • Not suitable for CI • Not DevOps friendly Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions NFSROOT based Standard Pipeline • The image is usually generated from a "golden" node or by bootstrap. • The image is shared via NFS. • The OS boots from PXE • The root filesystem is mounted from a NFS server File System Management Options • Read-Write: Each node needs its own OS file system • Read-Only: Potential high memory footprint Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions NFSROOT based Pros • Easy to apply changes online • Easy to manage • Consistent configuration across the cluster • Still room for configuration managers Cons • NFS server becomes a SPOF • Limited scalability • Potential large memory footprint and hard to maintain rwtab file (read-only option) • Complex data structure and ongoing communication (read-write option) Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Stateless with ramdisk Standard Pipeline • The image is usually generated from a "golden" node. • The image is uploaded to a TFTP/HTTP/FTP/NFS server. • The OS boots from PXE • The root file system is loaded in the memory of the system. • Mutable files will use additional memory. Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Stateless with ramdisk Pros • Resilient to TFTP/HTTP/FTP/NFS failures • Resilient to local disk failures • Easy to manage • Consistent configuration across the cluster Cons • Usually it generates a massive ramdisk (500M to 4GB) • Limited scalability with single server • Long term boot process if the image is distributed through 1GB eth • Large (sometimes huge) memory footprint • ramdisk supports only 16 bits Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Semi State-less / State-lite with ramdisk Standard Pipeline • The image is usually generated from a "golden" node. • The image is uploaded to a TFTP/HTTP/FTP/NFS server. • The OS boots from PXE. • The root file system is loaded in the ram of the system. • The rootfs is partially located in a NFS server or cluster file system. Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Semi State-less / State-lite with ramdisk Pros • Resilient to TFTP/HTTP/FTP/NFS failures • Resilient to local disk failures • Less memory footprint • Allows applying (some) changes online in the portion of the file system located in the shared FS. Cons • NFS is a SPOF. Cluster FS without HA is a SPOF • Limited scalability with single TFTP/HTTP/FTP/NFS server • ramdisk supports only 16 bits • Complex maintenance and administration • Not suitable for CI • Not DevOps friendly Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Improvements in diskless provisioning Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Improvements in diskless provisioning NFS-root and state-lite solutions have limited scalability and SPOF. By booting the OS with a native cluster file system client we achieved: Pros • Great scalability • RDMA support • Fully resilient solution • Consistent configuration across the cluster • Small memory footprint Cons • Requires maintaining /etc/rwtab • Changes which require updating the rwtab file, also require a reboot Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Read-only BeeGFS root file system DEMO Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Read-only Lustre root file system Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Comparison Provisioning Mechanism Memory (MB) Data Transf. (MB/node) Deployment 285 400 read-only NFS root 300 42 read-only BeeGFS root 320 44 read-only Lustre root 330 44 Table: Statistics based on sNow! CentOS 7.3 minimal template (611 packages, 1638MB of file system). Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Introducing OverlayFS root Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions About OverlayFS • Developed by Miklos Szeredi • implements a union mount for other file systems. • merged into the Linux kernel mainline in 2014, in kernel version 3.18. • OverlayFS supports whiteouts and opaque directories in the upper file system to allow file and directory deletion. • improved in version 4.0 (fixed issues related with inode utilization and memory footprint) Jordi Blasco (HPCNow!) Modernising HPC cluster provisioning Motivation Deployment Provisioning Image Provisioning Improvements in diskless provisioning Introducing OverlayFS root Conclusions Architecture of sNow! OverlayFS root Overlay mount options: mount -t overlay overlay \ -o lowerdir=/lower,upperdir=/upper,workdir=/work /merged
Recommended publications
  • Lecture 9: Data Storage and IO Models Lecture 9
    Lecture 9 Lecture 9: Data Storage and IO Models Lecture 9 Announcements • Submission Project Part 1 tonight • Instructions on Piazza! • PS2 due on Friday at 11:59 pm • Questions? Easier than PS1. • Badgers Rule! Lecture 9 Today’s Lecture 1. Data Storage 2. Disk and Files 3. Buffer Manager - Prelims 3 Lecture 9 > Section 1 1. Data Storage 4 Lecture 9 > Section 1 What you will learn about in this section 1. Life cycle of a query 2. Architecture of a DBMS 3. Memory Hierarchy 5 Lecture 9 > Section 1 Life cycle of a query Query Result Query Database Server Query Execute Parser Optimizer Select R.text from |…|……|………..|………..| Report R, Weather W |…|……|………..|………..| where W.image.rain() Scheduler Operators |…|……|………..|………..| and W.city = R.city |…|……|………..|………..| and W.date = R.date |…|……|………..|………..| and |…|……|………..|………..| R.text. |…|……|………..|………..| matches(“insurance claims”) |…|……|………..|………..| |…|……|………..|………..| |…|……|………..|………..| |…|……|………..|………..| Query Query Syntax Tree Query Plan Result Segments 6 Lecture 9 > Section 1 > Architecture of a DBMS Internal Architecture of a DBMS query Query Execution data access Storage Manager I/O access 7 Lecture 9 > Section 1 > Storage Manager Architecture of a Storage Manager Access Methods Sorted File Hash Index B+-tree Heap File Index Manager Buffer Manager Recovery Manager Concurrency Control I/O Manager IO Accesses In Systems, IO cost matters a ton! 8 Lecture 9 > Section 1 > Data Storage Data Storage • How does a DBMS store and access data? • main memory (fast, temporary) • disk (slow, permanent) • How
    [Show full text]
  • Rootless Containers with Podman and Fuse-Overlayfs
    CernVM Workshop 2019 (4th June 2019) Rootless containers with Podman and fuse-overlayfs Giuseppe Scrivano @gscrivano Introduction 2 Rootless Containers • “Rootless containers refers to the ability for an unprivileged user (i.e. non-root user) to create, run and otherwise manage containers.” (https://rootlesscontaine.rs/ ) • Not just about running the container payload as an unprivileged user • Container runtime runs also as an unprivileged user 3 Don’t confuse with... • sudo podman run --user foo – Executes the process in the container as non-root – Podman and the OCI runtime still running as root • USER instruction in Dockerfile – same as above – Notably you can’t RUN dnf install ... 4 Don’t confuse with... • podman run --uidmap – Execute containers as a non-root user, using user namespaces – Most similar to rootless containers, but still requires podman and runc to run as root 5 Motivation of Rootless Containers • To mitigate potential vulnerability of container runtimes • To allow users of shared machines (e.g. HPC) to run containers without the risk of breaking other users environments • To isolate nested containers 6 Caveat: Not a panacea • Although rootless containers could mitigate these vulnerabilities, it is not a panacea , especially it is powerless against kernel (and hardware) vulnerabilities – CVE 2013-1858, CVE-2015-1328, CVE-2018-18955 • Castle approach : it should be used in conjunction with other security layers such as seccomp and SELinux 7 Podman 8 Rootless Podman Podman is a daemon-less alternative to Docker • $ alias
    [Show full text]
  • System Calls System Calls
    System calls We will investigate several issues related to system calls. Read chapter 12 of the book Linux system call categories file management process management error handling note that these categories are loosely defined and much is behind included, e.g. communication. Why? 1 System calls File management system call hierarchy you may not see some topics as part of “file management”, e.g., sockets 2 System calls Process management system call hierarchy 3 System calls Error handling hierarchy 4 Error Handling Anything can fail! System calls are no exception Try to read a file that does not exist! Error number: errno every process contains a global variable errno errno is set to 0 when process is created when error occurs errno is set to a specific code associated with the error cause trying to open file that does not exist sets errno to 2 5 Error Handling error constants are defined in errno.h here are the first few of errno.h on OS X 10.6.4 #define EPERM 1 /* Operation not permitted */ #define ENOENT 2 /* No such file or directory */ #define ESRCH 3 /* No such process */ #define EINTR 4 /* Interrupted system call */ #define EIO 5 /* Input/output error */ #define ENXIO 6 /* Device not configured */ #define E2BIG 7 /* Argument list too long */ #define ENOEXEC 8 /* Exec format error */ #define EBADF 9 /* Bad file descriptor */ #define ECHILD 10 /* No child processes */ #define EDEADLK 11 /* Resource deadlock avoided */ 6 Error Handling common mistake for displaying errno from Linux errno man page: 7 Error Handling Description of the perror () system call.
    [Show full text]
  • Ext4 File System and Crash Consistency
    1 Ext4 file system and crash consistency Changwoo Min 2 Summary of last lectures • Tools: building, exploring, and debugging Linux kernel • Core kernel infrastructure • Process management & scheduling • Interrupt & interrupt handler • Kernel synchronization • Memory management • Virtual file system • Page cache and page fault 3 Today: ext4 file system and crash consistency • File system in Linux kernel • Design considerations of a file system • History of file system • On-disk structure of Ext4 • File operations • Crash consistency 4 File system in Linux kernel User space application (ex: cp) User-space Syscalls: open, read, write, etc. Kernel-space VFS: Virtual File System Filesystems ext4 FAT32 JFFS2 Block layer Hardware Embedded Hard disk USB drive flash 5 What is a file system fundamentally? int main(int argc, char *argv[]) { int fd; char buffer[4096]; struct stat_buf; DIR *dir; struct dirent *entry; /* 1. Path name -> inode mapping */ fd = open("/home/lkp/hello.c" , O_RDONLY); /* 2. File offset -> disk block address mapping */ pread(fd, buffer, sizeof(buffer), 0); /* 3. File meta data operation */ fstat(fd, &stat_buf); printf("file size = %d\n", stat_buf.st_size); /* 4. Directory operation */ dir = opendir("/home"); entry = readdir(dir); printf("dir = %s\n", entry->d_name); return 0; } 6 Why do we care EXT4 file system? • Most widely-deployed file system • Default file system of major Linux distributions • File system used in Google data center • Default file system of Android kernel • Follows the traditional file system design 7 History of file system design 8 UFS (Unix File System) • The original UNIX file system • Design by Dennis Ritche and Ken Thompson (1974) • The first Linux file system (ext) and Minix FS has a similar layout 9 UFS (Unix File System) • Performance problem of UFS (and the first Linux file system) • Especially, long seek time between an inode and data block 10 FFS (Fast File System) • The file system of BSD UNIX • Designed by Marshall Kirk McKusick, et al.
    [Show full text]
  • File Handling in Python
    hapter C File Handling in 2 Python There are many ways of trying to understand programs. People often rely too much on one way, which is called "debugging" and consists of running a partly- understood program to see if it does what you expected. Another way, which ML advocates, is to install some means of understanding in the very programs themselves. — Robin Milner In this Chapter » Introduction to Files » Types of Files » Opening and Closing a 2.1 INTRODUCTION TO FILES Text File We have so far created programs in Python that » Writing to a Text File accept the input, manipulate it and display the » Reading from a Text File output. But that output is available only during » Setting Offsets in a File execution of the program and input is to be entered through the keyboard. This is because the » Creating and Traversing a variables used in a program have a lifetime that Text File lasts till the time the program is under execution. » The Pickle Module What if we want to store the data that were input as well as the generated output permanently so that we can reuse it later? Usually, organisations would want to permanently store information about employees, inventory, sales, etc. to avoid repetitive tasks of entering the same data. Hence, data are stored permanently on secondary storage devices for reusability. We store Python programs written in script mode with a .py extension. Each program is stored on the secondary device as a file. Likewise, the data entered, and the output can be stored permanently into a file.
    [Show full text]
  • In Search of the Ideal Storage Configuration for Docker Containers
    In Search of the Ideal Storage Configuration for Docker Containers Vasily Tarasov1, Lukas Rupprecht1, Dimitris Skourtis1, Amit Warke1, Dean Hildebrand1 Mohamed Mohamed1, Nagapramod Mandagere1, Wenji Li2, Raju Rangaswami3, Ming Zhao2 1IBM Research—Almaden 2Arizona State University 3Florida International University Abstract—Containers are a widely successful technology today every running container. This would cause a great burden on popularized by Docker. Containers improve system utilization by the I/O subsystem and make container start time unacceptably increasing workload density. Docker containers enable seamless high for many workloads. As a result, copy-on-write (CoW) deployment of workloads across development, test, and produc- tion environments. Docker’s unique approach to data manage- storage and storage snapshots are popularly used and images ment, which involves frequent snapshot creation and removal, are structured in layers. A layer consists of a set of files and presents a new set of exciting challenges for storage systems. At layers with the same content can be shared across images, the same time, storage management for Docker containers has reducing the amount of storage required to run containers. remained largely unexplored with a dizzying array of solution With Docker, one can choose Aufs [6], Overlay2 [7], choices and configuration options. In this paper we unravel the multi-faceted nature of Docker storage and demonstrate its Btrfs [8], or device-mapper (dm) [9] as storage drivers which impact on system and workload performance. As we uncover provide the required snapshotting and CoW capabilities for new properties of the popular Docker storage drivers, this is a images. None of these solutions, however, were designed with sobering reminder that widespread use of new technologies can Docker in mind and their effectiveness for Docker has not been often precede their careful evaluation.
    [Show full text]
  • System Calls and Standard I/O
    System Calls and Standard I/O Professor Jennifer Rexford http://www.cs.princeton.edu/~jrex 1 Goals of Today’s Class • System calls o How a user process contacts the Operating System o For advanced services that may require special privilege • Standard I/O library o Generic I/O support for C programs o A smart wrapper around I/O-related system calls o Stream concept, line-by-line input, formatted output, ... 2 1 System Calls 3 Communicating With the OS User Process signals systems calls Operating System • System call o Request to the operating system to perform a task o … that the process does not have permission to perform • Signal o Asynchronous notification sent to a process … to notify the process of an event that has occurred o 4 2 Processor Modes • The OS must restrict what a user process can do o What instructions can execute o What portions of the address space are accessible • Supervisor mode (or kernel mode) o Can execute any instructions in the instruction set – Including halting the processor, changing mode bit, initiating I/O o Can access any memory location in the system – Including code and data in the OS address space • User mode o Restricted capabilities – Cannot execute privileged instructions – Cannot directly reference code or data in OS address space o Any such attempt results in a fatal “protection fault” – Instead, access OS code and data indirectly via system calls 5 Main Categories of System Calls • File system o Low-level file I/O o E.g., creat, open, read, write, lseek, close • Multi-tasking mechanisms o Process
    [Show full text]
  • MINCS - the Container in the Shell (Script)
    MINCS - The Container in the Shell (script) - Masami Hiramatsu <[email protected]> Tech Lead, Linaro Ltd. Open Source Summit Japan 2017 LEADING COLLABORATION IN THE ARM ECOSYSTEM Who am I... Masami Hiramatsu - Linux kernel kprobes maintainer - Working for Linaro as a Tech Lead LEADING COLLABORATION IN THE ARM ECOSYSTEM Demo # minc top # minc -r /opt/debian/x86_64 # minc -r /opt/debian/arm64 --arch arm64 LEADING COLLABORATION IN THE ARM ECOSYSTEM What Is MINCS? My Personal Fun Project to learn how linux containers work :-) LEADING COLLABORATION IN THE ARM ECOSYSTEM What Is MINCS? Mini Container Shell Scripts (pronounced ‘minks’) - Container engine implementation using POSIX shell scripts - It is small (~60KB, ~2KLOC) (~20KB in minimum) - It can run on busybox - No architecture dependency (* except for qemu/um mode) - No need for special binaries (* except for libcap, just for capsh --exec) - Main Features - Namespaces (Mount, PID, User, UTS, Net*) - Cgroups (CPU, Memory) - Capabilities - Overlay filesystem - Qemu cross-arch/system emulation - User-mode-linux - Image importing from dockerhub And all are done by CLI commands :-) LEADING COLLABORATION IN THE ARM ECOSYSTEM Why Shell Script? That is my favorite language :-) - Easy to understand for *nix administrators - Just a bunch of commands - Easy to modify - Good for prototyping - Easy to deploy - No architecture dependencies - Very small - Able to run on busybox (+ libcap is perfect) LEADING COLLABORATION IN THE ARM ECOSYSTEM MINCS Use-Cases For Learning - Understand how containers work For Development - Prepare isolated (cross-)build environment For Testing - Test new applications in isolated environment - Test new kernel features on qemu using local tools For products? - Maybe good for embedded devices which has small resources LEADING COLLABORATION IN THE ARM ECOSYSTEM What Is A Linux Container? There are many linux container engines - Docker, LXC, rkt, runc, ..
    [Show full text]
  • [13주차] Sysfs and Procfs
    1 7 Computer Core Practice1: Operating System Week13. sysfs and procfs Jhuyeong Jhin and Injung Hwang Embedded Software Lab. Embedded Software Lab. 2 sysfs 7 • A pseudo file system provided by the Linux kernel. • sysfs exports information about various kernel subsystems, HW devices, and associated device drivers to user space through virtual files. • The mount point of sysfs is usually /sys. • sysfs abstrains devices or kernel subsystems as a kobject. Embedded Software Lab. 3 How to create a file in /sys 7 1. Create and add kobject to the sysfs 2. Declare a variable and struct kobj_attribute – When you declare the kobj_attribute, you should implement the functions “show” and “store” for reading and writing from/to the variable. – One variable is one attribute 3. Create a directory in the sysfs – The directory have attributes as files • When the creation of the directory is completed, the directory and files(attributes) appear in /sys. • Reference: ${KERNEL_SRC_DIR}/include/linux/sysfs.h ${KERNEL_SRC_DIR}/fs/sysfs/* • Example : ${KERNEL_SRC_DIR}/kernel/ksysfs.c Embedded Software Lab. 4 procfs 7 • A special filesystem in Unix-like operating systems. • procfs presents information about processes and other system information in a hierarchical file-like structure. • Typically, it is mapped to a mount point named /proc at boot time. • procfs acts as an interface to internal data structures in the kernel. The process IDs of all processes in the system • Kernel provides a set of functions which are designed to make the operations for the file in /proc : “seq_file interface”. – We will create a file in procfs and print some data from data structure by using this interface.
    [Show full text]
  • USB Composite Gadget Using CONFIG-FS on Dra7xx Devices
    Application Report SPRACB5–September 2017 USB Composite Gadget Using CONFIG-FS on DRA7xx Devices RaviB ABSTRACT This application note explains how to create a USB composite gadget, network control model (NCM) and abstract control model (ACM) from the user space using Linux® CONFIG-FS on the DRA7xx platform. Contents 1 Introduction ................................................................................................................... 2 2 USB Composite Gadget Using CONFIG-FS ............................................................................. 3 3 Creating Composite Gadget From User Space.......................................................................... 4 4 References ................................................................................................................... 8 List of Figures 1 Block Diagram of USB Composite Gadget............................................................................... 3 2 Selection of CONFIGFS Through menuconfig........................................................................... 4 3 Select USB Configuration Through menuconfig......................................................................... 4 4 Composite Gadget Configuration Items as Files and Directories ..................................................... 5 5 VID, PID, and Manufacturer String Configuration ....................................................................... 6 6 Kernel Logs Show Enumeration of USB Composite Gadget by Host ................................................ 6 7 Ping
    [Show full text]
  • High Velocity Kernel File Systems with Bento
    High Velocity Kernel File Systems with Bento Samantha Miller, Kaiyuan Zhang, Mengqi Chen, and Ryan Jennings, University of Washington; Ang Chen, Rice University; Danyang Zhuo, Duke University; Thomas Anderson, University of Washington https://www.usenix.org/conference/fast21/presentation/miller This paper is included in the Proceedings of the 19th USENIX Conference on File and Storage Technologies. February 23–25, 2021 978-1-939133-20-5 Open access to the Proceedings of the 19th USENIX Conference on File and Storage Technologies is sponsored by USENIX. High Velocity Kernel File Systems with Bento Samantha Miller Kaiyuan Zhang Mengqi Chen Ryan Jennings Ang Chen‡ Danyang Zhuo† Thomas Anderson University of Washington †Duke University ‡Rice University Abstract kernel-level debuggers and kernel testing frameworks makes this worse. The restricted and different kernel programming High development velocity is critical for modern systems. environment also limits the number of trained developers. This is especially true for Linux file systems which are seeing Finally, upgrading a kernel module requires either rebooting increased pressure from new storage devices and new demands the machine or restarting the relevant module, either way on storage systems. However, high velocity Linux kernel rendering the machine unavailable during the upgrade. In the development is challenging due to the ease of introducing cloud setting, this forces kernel upgrades to be batched to meet bugs, the difficulty of testing and debugging, and the lack of cloud-level availability goals. support for redeployment without service disruption. Existing Slow development cycles are a particular problem for file approaches to high-velocity development of file systems for systems.
    [Show full text]
  • Singularityce User Guide Release 3.8
    SingularityCE User Guide Release 3.8 SingularityCE Project Contributors Aug 16, 2021 CONTENTS 1 Getting Started & Background Information3 1.1 Introduction to SingularityCE......................................3 1.2 Quick Start................................................5 1.3 Security in SingularityCE........................................ 15 2 Building Containers 19 2.1 Build a Container............................................. 19 2.2 Definition Files.............................................. 24 2.3 Build Environment............................................ 35 2.4 Support for Docker and OCI....................................... 39 2.5 Fakeroot feature............................................. 79 3 Signing & Encryption 83 3.1 Signing and Verifying Containers.................................... 83 3.2 Key commands.............................................. 88 3.3 Encrypted Containers.......................................... 90 4 Sharing & Online Services 95 4.1 Remote Endpoints............................................ 95 4.2 Cloud Library.............................................. 103 5 Advanced Usage 109 5.1 Bind Paths and Mounts.......................................... 109 5.2 Persistent Overlays............................................ 115 5.3 Running Services............................................. 118 5.4 Environment and Metadata........................................ 129 5.5 OCI Runtime Support.......................................... 140 5.6 Plugins.................................................
    [Show full text]