Application Performance and Flexibility on ExoKernel Systems

Editor's Notes

• #2: Hello good morning. We are going to present the research paper on “Application Performance and Flexibility on Exokernel Systems” by the mentioned authors.
• #3: This will be the outline of the presentation.
• #4: An operating system is the interface between applications and physical resources of the computer. Traditional operating systems have a large kernel that creates high-level abstractions of these resources. The kernel is a fundamental part of a modern computer's operating system. Responsibilities : Security Management of resources Enabling a uniform execution environment
• #5: Monolithic kernels In a monolithic kernel, all OS services run along with the main kernel thread, thus also residing in the same memory area. This approach provides rich and powerful hardware access. it is "easier to implement a monolithic kernel"[29] than microkernels. Microkernels Microkernel is the term describing an approach to Operating System design by which the functionality of the system is moved out of the traditional "kernel", into a set of "servers" that communicate through a "minimal" kernel, leaving as little as possible in "system space" and as much as possible in "user space". Hybrid (or Modular) kernels Hybrid kernels are used in most commercial operating systems such as Microsoft Windows NT 3.1, NT 3.5, NT 3.51, NT 4.0, 2000, XP, Vista, 7, 8, and 8.1 . Apple Inc's own Mac OS X uses a hybrid kernel called XNU which is based upon code from Carnegie Mellon's Mach kernel and FreeBSD's monolithic kernel. Nanokernels A nanokernel delegates virtually all services — including even the most basic ones like interrupt controllers or the timer — to device drivers to make the kernel memory requirement even smaller than a traditional microkernel.[38] Exokernels Exokernels are a still experimental approach to operating system design. They differ from the other types of kernels in that their functionality is limited to the protection and multiplexing of the raw hardware, providing no hardware abstractions on top of which to develop applications. This separation of hardware protection from hardware management enables application developers to determine how to make the most efficient use of the available hardware for each specific program. ------------------------------------------------------------------------------------ Monolithic kernels In a monolithic kernel, all OS services run along with the main kernel thread, thus also residing in the same memory area. This approach provides rich and powerful hardware access. Some developers, such as UNIX developer Ken Thompson, maintain that it is "easier to implement a monolithic kernel"[29] than microkernels. The main disadvantages of monolithic kernels are the dependencies between system components — a bug in a device driver might crash the entire system — and the fact that large kernels can become very difficult to maintain. Microkernels Microkernel (also abbreviated μK or uK) is the term describing an approach to Operating System design by which the functionality of the system is moved out of the traditional "kernel", into a set of "servers" that communicate through a "minimal" kernel, leaving as little as possible in "system space" and as much as possible in "user space". A microkernel that is designed for a specific platform or device is only ever going to have what it needs to operate. The microkernel approach consists of defining a simple abstraction over the hardware, with a set of primitives or system calls to implement minimal OS services such as memory management, multitasking, and inter-process communication. Hybrid (or Modular) kernels Hybrid kernels are used in most commercial operating systems such as Microsoft Windows NT 3.1, NT 3.5, NT 3.51, NT 4.0, 2000, XP, Vista, 7, 8, and 8.1 . Apple Inc's own Mac OS X uses a hybrid kernel called XNU which is based upon code from Carnegie Mellon's Mach kernel and FreeBSD's monolithic kernel. Nanokernels A nanokernel delegates virtually all services — including even the most basic ones like interrupt controllers or the timer — to device drivers to make the kernel memory requirement even smaller than a traditional microkernel.[38] Exokernels Exokernels are a still experimental approach to operating system design. They differ from the other types of kernels in that their functionality is limited to the protection and multiplexing of the raw hardware, providing no hardware abstractions on top of which to develop applications. This separation of hardware protection from hardware management enables application developers to determine how to make the most efficient use of the available hardware for each specific program.
• #6: Fixed, high-level abstractions degrade application performance. Because there is no single way to implement an abstraction that is best for all applications. These abstractions hide information about physical resources which applications execute. It is just like a virtual machine.   This design is inefficient for three reasons: it denies applications the advantages of domain-specific optimizations, it discourages changes to the implementations of existing abstractions, and it restricts the flexibility of applications builders because new abstractions can only be added on top of existing ones   Because of these reasons, we encounter performance issues.   An analogy can be found between Reduced Instruction Set (RISC) processors, which are faster than the more complicated higher-level CISC (Complex instruction set computing) instructions.   Another analogy is that between low-level Assembly code compared to high-level programming languages like C++ or Java.
• #7: The exokernel operating system architecture extends the traditional hardware interface by granting application-level management of low-level physical resources.   As a result, performance, flexibility, and functionality of applications are improved significantly.   “Applications know better than operating systems what the goal of their resource management decisions should be and therefore, they should be given as much control as possible over those decisions” (Engler, 1995).   The job of the exokernel is to securely multiplex all available hardware while maintaining protection of these resources at a fine-grained level.   Library operating systems (LibOSes) work on top of the exokernel interface, importing kernel information and data structures as needed for memory management, process scheduling, I/O. Application programmers choose only those library operating systems that they want so that they are able to implement abstractions in the way that they want to, and they only import those libraries that they are going to use.
• #8: Just like exokernels, to get the maximum use of the human body it accesses the human body directly
• #9: Can applications actually achieve significant performance improvements on an exokernel? Will traditional applications (unaltered UNIX applications) have a performance hit? Is global performance compromised when no centralized authority decides scheduling and multiplexing policies? Does the lack of a centralized management policy for shared OS structures lower the integrity of the system?
• #10: Exokernels should avoid resource management; they should only manage resources to the extent that require protection. Explicit resource allocation Physical names capture useful information Expose revocation policies to all the applications Expose all the system inforamtion ------------------------------------------------------------------   Exokernels must give library operating systems maximum freedom in managing physical resources while protecting them from each other   To do this, an exokernel performs three tasks: it tracks ownership of resources, ensures protection by guarding all resource usage or binding points revokes access to resources.
• #12: Multiplexing the CPU   The kernel multiplexes the CPU by dividing it into time slices of a fixed number of ticks. An environment may then allocate time slices in order to be scheduled on the processor. This allows a process to save or restore its own registers.   Multiplexing Main Memory   Multiplexing main memory is a rather straightforward process. It only has to guard memory accesses between competing libOSes rather than from within a single libOS.   Multiplexing the Stable Storage   Xok uses XN, its native file system, to provide libOSes with access to stable storage at the level of disk blocks, To determining the access rights of a requesting process to a given disk block as efficiently as possible.
• #13: LibOSes have one or more libFSes. An exokernel must provide a means to safely multiplex disks among multiple library file systems (libFSes). Creating new file formats should be simple and lightweight. allow multiple libFSes to safely share files at the raw disk block and metadata level. Storage system must be efficient. Storage system should facilitate cache sharing among libFSes allow them to easily address problems of cache coherence, security, and concurrency.  Designing a flexible exokernel stable storage system has proven difficult: XN is our fourth design.
• #14: One such file system written to multiplex the stable storage is XN. Employs UDFs (untrusted deterministic functions). Since we are accessing low level abstractions,; UDFs are metadata translation functions specific to each file type. UDFs allow the kernel to safely and efficiently handle any metadata layout without understanding the layout itself. The most difficult requirement for XN is efficiently determining the access rights of a given principal to a given disk block. CFFS: A library file system (collocating fast file system)  A UNIX like library file system with special reference to additional protection guarantees it provides.