Shulou(Shulou.com)06/01 Report--

This article will explain in detail what improvements have been made in RHEL 6 kernel optimization. The quality of the article is high, so Xiaobian shares it with you for reference. I hope you have a certain understanding of relevant knowledge after reading this article.

In RHEL 6, Red Hat will only provide support for kernel virtual machines (KVM). KVM exists as a single module in the kernel, meaning that KVM virtual machines run as processes on top of that module. This makes virtual machines much easier to manage. RHEL 6 makes some kernel improvements to optimize Linux virtualization services. A brief description of some of the most important improvements.

KVM virtualization technology provides RHEL 6 with a next-generation virtualization solution. Because KVM is more efficient than Xen updates, Red Hat will abandon Xen altogether in RHEL 6 (but will provide a way to port Xen virtual machines).

Optimized process handling

RHEL processes can be organized in buckets: entities that can allocate CPU and other resources. Because virtual machines also run as processes, this property also becomes part of the virtual machine runtime mechanism.

In order for virtual machines to run smoothly in KVM, each virtual CPU is processed as a thread by the scheduler. For efficient scheduling, the kernel enables a new mode of operation called guest mode. A virtual guest cannot make system calls directly into the Linux kernel (as a process in system mode can), but it can use administrative calls to talk to the hypervisor. This new mode of operation takes advantage of new Linux kernel features such as scheduling, counting, and kernel same-page merging (KSM).

Hardware support improvements

Kernel upgrades are only part of KVM performance. The other part is the hardware itself. The RHEL 6 kernel takes advantage of several features that are implemented entirely in hardware. First, regarding CPU support, there is Extended Page Table (EPT). This feature reduces the need for emulation, which allows virtualized services to run faster. This is followed by what is known as IOMMU in AMD environments and VT-d in Intel environments. This feature allows users to safely use physical I/O devices directly while protecting the same devices from accidental use by other users. This protection is necessary because if a user writes data to an occupied device, it could cause the host to crash.

Another important feature is SR-IOV. This feature refers to the virtual I/O interface on the PCI bus, securely sharing real hardware. This feature is especially important for network adapters, which allow a physical device to simultaneously provide virtual devices that can be used by multiple guest users. *** An important hardware feature that can be used in a virtual environment is NPIV. This feature allows sharing of storage devices, which means that each user can use a portion of the storage device.

To take advantage of these hardware improvements, the RHEL kernel has been enhanced. For example, CPU enhancements can allow users to allocate more than 64 CPUs. Furthermore, due to the "locking" nature of the kernel RCU synchronization mechanism-virtually no locks, but a mechanism to avoid locks and improve performance in SMP symmetric multiprocessor environments-there is a minimum number of CPUs that can be allocated.

RHEL 6 also has some memory enhancements. First of all, *** item enhancement is transparent large memory paging. This means that large memory paging (which allows the kernel to allocate large blocks of memory rather than small 4KB blocks) is dynamic, so there is no paging plan for memory at system startup. Another feature that benefits KVM is kernel same-page merging. This feature allows multiple virtual machines to access the same administrative page simultaneously, clearing all memory pages during startup and improving the performance of Windows virtual machines.

The next improvement is block I/O. The first is local asynchronous IO and preadv/pwritev system calls. With these, you can group the memory areas you need before reading/writing, improving performance in multithreaded environments. The kernel also now has MSI interrupt support. MSI interrupts are tied to PCI devices, allowing you to work on multiple parallel interrupt lines. Another related improvement is block alignment, which is now optimized from its default values for better performance. All of these improvements are designed to provide kernel-level *** performance.

Special I/O enhancements also benefit network interface performance. One of the most important enhancements is the virtual hosting network that moves a portion of network space from user space to kernel space. This is useful for virtio drivers. Virtion drivers also provide KVM users with better network performance while allowing users to use TAP drivers to talk to virtual appliances. The advantage of using virtual host nets is that they anchor directly into the kernel and thus provide much better performance than the qemu approach, which requires emulation.

About RHEL 6 kernel optimization what improvements are shared here, I hope the above content can be of some help to everyone, you can learn more knowledge. If you think the article is good, you can share it so that more people can see it.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.