Red Hat Releases RHEL 6.2

New release of Red Hats flagship Linux is more secure, flexibile and faster than ever

Red Hat has released a new version of their flagship Enterprise Linux OS.  RHEL 6.2 consolidates patches since REHL 6.1 plus offers some substantial improvements in security, virtualization, performance and flexibility. I spoke with Tim Burke, VP of Linux engineering at Red Hat about the new release.

It has been about a year since Red Hat released version 6.0 of REHL. They usually release an update every 6 months or so, so this 6.2 release is right on schedule. This certainty and stability around release cycles is something that Red Hat's customers have come to rely on. Burke says this is one of the things that separates Red Hat from other Linux companies.  They have a stable model that enterprises can plan around.

The new 6.2 release of course consolidates all updates since RHEL 6.1.  But that is not all that the Red Hat engineers have packed into 6.2. According to Red Hat some of the new features and functionality include:

1) Performance - Red Hat Enterprise Linux 6.2 continues enhancements to accelerate I/O throughput, increasing as much as 30% in this release.  Red Hat Enterprise Linux has enhanced file system functionality that reduces read-write times which boosts overall system utilization.

The latest SAP Standard Application Benchmarks illustrate that Red Hat Enterprise Linux 6 supports more than 22,713 users on a single system.   This is the largest Linux result submitted to SAP to-date.(Certificate number 22713, SAP enhancement package 4 for SAP ERP 6.0)  For more information on this benchmark,  visitwww.sap.com/benchmark

2) Virtualization around KVM:

• Virtual CPU timeslice sharing for multi-processor guests is a new feature in RHEL 6.2. Scheduler changes within the kernel now allow for virtual CPUs inside a guest to make more efficient use of the timeslice allocated to the guest before processor time is yielded back to the host. This change is especially beneficial to large SMP systems that have traditionally experienced guest performance lag due to inherent lock holder preemption issues.  In summary, this new feature eliminates resource consuming system overhead so that a guest can use more of the CPU resources assigned to them much more efficiently.

• KVM network performance is a critical requirement for Virtualization and cloud based products and solutions.  Improvements have been made in KVM to more efficiently process small (<4kb) messages.  In addition, network drivers for KVM (virtio-net) have been enhanced.  Social networking, which generates a large number of small-sized messages, can now be accelerated and realize better performance with the enhancements made in KVM functionality. 

• Guest memory pinning KVM enhancements.  The libvirt management API now includes tuning improvements that result in a better out-of-the-box experience on NUMA environments. Prior to this enhancement, this tuning would have had to be done at a lower level and required a deeper understanding of NUMA configurations using libNUMA. It is also now possible to pin memory associated with the NUMA node at the time of guest creation.  This is in addition to the ability to pin specific CPUs to guests which is available in prior releases.

• A Guest debugging feature results in a simpler and easier administrative method of debugging virtualized guests.  The process improvement is  via libvirt API and the virsh command line tool.  It is now possible to have a trigger for interrupts or initiate a back trace as a means to debug guest environments

Access to pre-boot guest environment is now possible by allowing administrators to monitor and audit information that outputs by the virtual machine BIOS and kernel. This allows interactive use of the virsh console prior to  guest boot up. A new component called sgabios has also been introduced to support this functionality.  Benefit of this functionality enables Administrators to be able to ensure the integrity and settings of virtualized guests before they are created.  This helps audit and closely monitor all guests.

3) Cgroups

• cgroups (control groups) is a function introduced in Red Hat Enterprise Linux 6 that can limit, account for and isolate resource usage (CPU, memory, disk I/O, etc.) of process groups. Many improvements and new functionality was introduced with Red Hat Enterprise Linux 6.2.; Fine-grained control of servers, make its possible to achieve higher system utilization.

• cgroup processor ceilings gives customers the ability to efficiently manage SLA’s, in particular, and provides customers with the ability to establish service guarantees, similar to those associated with network Quality of Service (QoS). For service providers or internal IT organizations who deliver applications or hosted services via multi-tenant environments, can now provide the ability to set maximum level of CPU consumption associated with a given application, business process or a virtual machine.

4) Hardware Support

Red Hat continues to embark on co-development relationships with the leading hardware vendors so that the latest advancement are included in our releases. There are number of RAS features to support our microprocessor partners and better performance optimization that avoids unnecessary cache delays.

Server IO continues to improve with greater bandwidth.  In Red Hat Enterprise Linux 6.2, support for PCI-e 3.0 as well as USB 3.0 gives customers the ability to realize significant improvements in faster and wider buses and provides system manufacturers and I/O developers a solid foundation on Red Hat for them to rapidly develop new and leading edge I/O devices for the global market.

5) High Availability

When an enterprise deploys their applications to run in a Red Hat Enterprise Linux 6.2 guest hosted by VMware,  applications can be clustered for High Availability (HA).  For virtual machines, this release includes full support for use of GFS2 which is a shared disk file system for Linux system clusters.  The result is additional deployment flexibility for those customer who require HA within a portion of their virtualized environment, as well as full support for Red Hat Enterprise Linux on the VMware hypervisor.

Simplified, intuitive User Interface (UI) for cluster configuration has been updated to include Role-Based Access Control (RBAC), which enables fine-grained access levels by defining specific user classes to have access to specific clusters. 

UDP-Unicast support improvements makes it easier and faster to set up a clustering infrastructure.  UDP-unicast in contrast to IP multitasking, offers a similar approach to cluster configuration and is an established protocol for cluster communication.

6) Storage and File Systems

Red Hat Enterprise Linux 6.2 adds enhancements to file system and storage functionality.  Full support of iSCSI extension for RDMA enables even the most demanding storage environments to realize the benefits of low latency and high throughput through a standard SAN implementation based on 10 Gb Ethernet.  This improvement can remove the need for expensive or other dedicated interconnect fabrics. 

Improved large data transfer rates with support for pNFS allows clients to access storage devices directly and in parallel. PNFS is an extension of NFS that provides significantly larger data transfer rates compared to transition NFS architectures.

XFS has been enhanced to support delayed logging of meta data and significantly improves performance for highly paralleled meta-data intensive workloads, such as thousands of small files in a directory as found in today's workload intensive social networking environments.

It is now possible to run clustered Samba services in an active-active setup using a shared GFS2 file system on the Red Hat cluster infrastructure.  Supporting multiple active instances of Samba in a cluster will not only improve overall throughput for Samba services, but also increases the overall availability for a large Samba environment. 

7) Networking

Networking traffic can sometimes be slowed down due to lack of cache efficiency when multiple CPU’s are involved.  Red Hat Enterprise Linux 6.2 provides customers the ability to dedicate CPU resources for managing networking traffic, which results in improved throughput while providing finer grain control over more resources.  Significant improvements in throughput can be realized as much as a 30% increase.

Stream Control Transmission Protocol (SCTP) support separates messages and control of the information into separate streams, facilitating a less congested streaming of related content, such as video and voice.  The result is lower latency while also enhancing the security of transmitted information. In addition, SCTP supports the rerouting of streaming content to another network connection if a failed network occurs, without interruption to the application. 

8) Identity Management

A new feature referred to as Identity Management in Red Hat Enterprise Linux 6.2 provides the administrative tools to quickly install, configure and manage server authentication and authorization in Linux enterprise environments.  This feature provides the option to interoperate with Microsoft Active Directory.  Enterprises can now manage Linux environments easily and cost-effectively.  Centralized identity management and host-based access control reduces administrative overhead, streamlines provisioning and improves overall security.

As can be seen from the above, to call this latest version of RHEL a minor release does not quite do it justice.  There is a lot here. If you are already a RHEL customer, there really is no reason not to upgrade. If you are considering going Linux, RHEL can deliver the enterprise scalabity and functionality you need.

Copyright © 2011 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022