Recently when benchmarking the Btrfs and EXT4 file-systems we were left surprised that the performance of the next-generation Btrfs file-system had regressed against EXT4 to the point where the evolutionary file-system is measurably faster in a greater number of disk benchmarks. In fact, even with solid-state drives and Btrfs offering an SSD optimized mode, it still conceded to EXT4. It turns out that in the Linux 2.6.35 kernel, Btrfs regressed. This regression should have been fixed with the Linux 2.6.36 kernel, but recently when benchmarking EXT4/Btrfs against ZFS-FUSE on a 2.6.36 development snapshot we found its performance to still be poor for Btrfs compared to EXT4. To confirm where these two most prominent Linux file-systems are at right now, we have new EXT4 and Btrfs performance results from the Linux 2.6.34, 2.6.35, and 2.6.36-rc3 kernels.
The Q3'2010 update to the Phoronix Test Suite introduces new test profiles, provides new analytics capabilities, supports testing under a more diverse selection of hardware and software, and provides numerous other features for those looking to deploy this leading automated testing platform within enterprise environments.
Last week we reported that a native ZFS implementation for Linux is soon being released that is based upon the work by Lawrence Livermore National Laboratory to bring Sun's ZFS file-system to Linux as a CDDL-licensed kernel module. As said though in that article, there is already a ZFS module for FUSE (File-system in User-space) that is already available and with it not living in the GPL-land of the Linux kernel, it is legally allowed, but it does not come without some performance overhead. Over the weekend though there's been some discussions in the related forum thread and elsewhere about the dependability of ZFS-FUSE and what the level of impact on using FUSE really amounts to in real-world usage. We have tested the ZFS-FUSE -- both the latest stable and Git snapshots -- and have compared this alternate ZFS Linux implementation to that of the native EXT4 and Btrfs.
Earlier this month we delivered benchmarks comparing the ZFS, EXT4, and Btrfs file-systems from both solid-state drives and hard drives. The EXT4 file-system was the clear winner in terms of the overall disk performance while Btrfs came in second followed by Sun's ZFS in FreeBSD 8.2. It was a surprise that in our most recent testing the EXT4 file-system turned around and did better than the next-generation Btrfs file-system, but it turns out that Btrfs regressed hard in Linux 2.6.35 as to be found in Ubuntu 10.10 and other soon-to-be-released distributions. However, regardless of where Btrfs is performing, its speed can be boosted by enabling its transparent zlib compression support.
Prior to the emergence of Btrfs as a viable next-generation Linux file-system, Sun's ZFS file-system was sought after for Linux due to its advanced feature-set and capabilities compared to EXT3 and other open-source file-systems at the time. While ZFS support has worked its way into OpenSolaris, FreeBSD, NetBSD, and other operating systems, ZFS had not been ported to Linux as its source-code is distributed under the CDDL license, which is incompatible with the GNU GPL barring it from integration into the mainline Linux kernel. Next month, however, a working ZFS module for the Linux kernel without a dependence on FUSE will be publicly released.
While we benchmark the latest Linux kernel code on a daily basis at kernel-tracker.phoromatic.com using our automated testing platform built on the Phoronix Test Suite, now that the Linux 2.6.35 kernel was released, we have run a formalized set of kernel benchmarks on a ThinkPad W510 notebook with an Intel Core i7 CPU to see how the Linux 64-bit kernel is running with this high-end notebook under the Linux 2.6.32, 2.6.33, 2.6.34, and 2.6.35 releases.
ZFS is often looked upon as an advanced, superior file-system and one of the strong points of the Solaris/OpenSolaris platform while most feel that only recently has Linux been able to catch-up on the file-system front with EXT4 and the still-experimental Btrfs. ZFS is copy-on-write, self-healing with 256-bit checksums, supports compression, online pool growth, scales much better than the UFS file-system commonly used on BSD operating systems, supports snapshots, supports deduplication, and the list goes on for the features of this file-system developed by Sun Microsystems. In this article we are seeing how well the performance of the ZFS file-system under PC-BSD/FreeBSD 8.1 stacks up to UFS (including UFS+J and UFS+S) and on the Linux side with EXT4 and Btrfs.
As was mentioned in last Friday's article, Which Is Faster: Debian Linux or FreeBSD, tests of FreeBSD atop the ZFS file-system (rather than UFS2+S) are currently underway and those results are expected to be published in full later this week as the ZFS disk performance is compared directly to UFS2+S, UFS2+J, and also Ubuntu Linux with the EXT4 and Btrfs file-systems. Today though we have a few ZFS performance numbers to share as we look at the performance of the new CAM-ATA sub-system on FreeBSD.
In previous articles I have hinted that at Phoronix we are working to take advantage of the Btrfs file-system within the Phoronix Test Suite and Phoromatic to provide an interesting feature that will further expand our automated testing capabilities, but how does this file-system come into play? Well, here is what's being worked on and it should be of terrific value to many people.
Three years ago Intel had released PowerTop, an open-source utility for Linux that would analyze how well your laptop was conserving power and would allow users to easily tune their system for maximum battery life via simple power optimizations. By simply running this utility, some users were able to significantly extend their battery life. However, is this utility still useful and needed with a modern Linux desktop? The most recent release of PowerTop (v1.11) was a year and a half ago, so we are seeing how well PowerTop is still able to reduce the power consumption of Intel notebooks/netbooks running Linux.
Yesterday we reported that Ubuntu 10.10 gained Btrfs installation support and since then we have been trying out this Btrfs support in Ubuntu "Maverick Meerkat" and have a fresh set of Btrfs benchmarks to serve up.
Last month we looked at the cost of running Compiz by means of looking at how the window manager affected the frame-rate of several different games and whether compositing was used. We also tested out several different drivers and pieces of hardware. When Compiz was running rather than GNOME's Metacity it often caused a measurable drop in the OpenGL performance and then we later found this to be the case too with KDE's KWin. Today we are seeing if and how using Mutter, the window manager for the GNOME 3.0 desktop that uses Clutter-based compositing, will affect the performance of several different open-source games.
With MeeGo using Btrfs by default, Canonical making plans for Btrfs in as soon as Ubuntu 10.10, and Novell now pushing Btrfs in openSUSE, among other milestones for this advanced Linux file-system, we decided to see where the Btrfs performance is now at with the Linux 2.6.35 kernel that's currently in development. We compare the Btrfs performance to EXT4 and see how some of the different mount options are affecting the file-system's performance in different benchmarks.
For the past six months we have been monitoring the performance of the very latest Linux kernel code on a daily basis across multiple systems. We have spotted a few regressions -- both positive and negative -- on occasion using our automated daily testing of the Linux kernel, but nothing like what we have encountered the past few days: the Linux 2.6.35 kernel performance has fallen hard. In fact, the performance has fallen very hard in a number of tests and right now, we would consider it a disaster. While the 2.6.35 code has not even seen its first release candidate yet, there are some massive performance drops in a variety of different tests that have yet to be corrected and nothing like we have encountered with previous kernel release cycles especially for a regression that has lived now for about one week.
With its continuing commitment to provide quarterly updates to its industry-leading automated testing and benchmarking platform, Phoronix Media has announced the immediate availability of Phoronix Test Suite 2.6 (codenamed "Lyngen").
Last week we published Arch Linux vs. Ubuntu benchmarks to finally lay to rest that for the overall system performance the speed of the rolling Arch Linux distribution is not too different from that of Ubuntu when running with similar package versions. One of the areas, however, where the performance was different with the "out of the box" experience was the OpenGL gaming where Ubuntu was using Compiz by default where as Arch had Metacity. This surprised many so we published another article entitled The Cost Of Running Compiz where we showed the performance penalties of a compositing window manager with different hardware and drivers. This led some to ask whether the performance of KWin also causes the OpenGL frame-rate to drop, so here are those KDE benchmarks.
Earlier this week we published benchmarks comparing Arch Linux and Ubuntu. There were only a few areas where the two Linux distributions actually performed differently with many of their core packages being similar, but one of the areas where the results were vastly different was with the OpenGL performance as Ubuntu uses Compiz by default (when a supported GPU driver is detected) where as Arch does not. This had surprised many within our forums so we decided to carry out a number of tests with different hardware and drivers to show off what the real performance cost is of running Compiz as a desktop compositing manager in different configurations.
The Kernel-based Virtual Machine (KVM) has been in the mainline Linux kernel since Linux 2.6.20 in early 2007 and over time it has become one of the most widely used virtualization platforms on Linux. Ubuntu uses KVM, Fedora uses KVM, and Red Hat Enterprise Linux has even switched from Xen to preferring KVM, among others. While occasionally we deliver new KVM virtualization benchmarks, we decided to investigate how the performance of KVM virtualization has changed -- if at all -- over the past two years for better or worse.
As we reported this morning, via a blog post and keynote to kick-off the start of the Ubuntu Developer Summit this week for engaging in Ubuntu 10.10 development activities, Mark Shuttleworth announced Ubuntu Light and Ubuntu Unity. Ubuntu Light is a new spin of Ubuntu that is being offered up to OEMs that are looking to offer Ubuntu Linux as part of a dual-boot installation on their PCs. Unity is the new Ubuntu desktop interface that is used by Ubuntu Light.
Last week in a FreeBSD status report we talked about the Chromium web-browser support on FreeBSD improving through a new subscription program whereby most of the FreeBSD patches are being kept closed-source for some length of time before being committed back upstream as open-source and reaching the hands of the non-subscribers. This caused some to question the work, but the developer behind this FreeBSD-Chromium subscription program, Sprewell, has written an editorial that we are now publishing. This details his beliefs concerning the future of open-source software business models.
With the recent release of GCC 4.5 and the forthcoming release of LLVM 2.7 that is expected in the coming days, we have decided to see how the performance of GCC compares to that of LLVM. For this testing we have results from GCC 4.3/4.4/4.5 compared to LLVM with its GCC front-end (LLVM-GCC) and against Clang, which is the native C/C++ compiler front-end for the Low-Level Virtual Machine.
Last week GCC 4.5.0 entered the world with improvements to the experimental C++0x support, Graphite-powered automatic parallelization support, compatibility with new ARM processors, Intel Atom and AMD Orochi optimizations, link-time optimization, and GCC plug-in support. Over the weekend we decided to benchmark this major update to the GNU Compiler Collection to see how its performance compares to that of GCC 4.3 and 4.4.
One of the benefits of Btrfs besides offering competitive performance against other Linux file-systems and SSD optimizations is its support for sub-volumes and writable snapshots. While Btrfs is still in development and is not yet used as a default file-system by any Linux distribution, Red Hat has been looking to capitalize upon the capabilities of Btrfs by introducing support for system rollbacks into Fedora. The Btrfs-based system rollback support has been a feature for Fedora 13 so with the release of the Fedora 13 Beta earlier this week we decided to further investigate this feature.
File-system benchmarks have become quite common to Phoronix in the age of EXT4 and Btrfs with these new file-systems driving much of the interest and as we have also been finding the Linux file-system performance to change between kernel releases (and in some cases, the performance has changed a great deal). Most recently we delivered benchmarks of EXT4 vs. Btrfs vs. Reiser4, but now a month later we are back with more Linux file-system benchmarks as we look to see if the disk performance has changed with the Linux 2.6.34 kernel.
Most of the time at Phoronix we focus on looking at the Linux graphics performance of the software drivers and hardware, since traditionally that has been one of the most troubling areas of Linux hardware support. Tides though have turned as AMD continues to back their own open-source strategy with providing documentation and pushing out code that enables open-source hardware support from 3D acceleration to power management, while Intel continues to back their fully open-source model too. Another area of hardware support that has caused much grief for users has been with printer support. Printers are not nearly as complex as a modern-day graphics processor, but the different vendors have not been quick to offer up any Linux support -- and binary-only drivers frequently back the ones that do. There is one printer manufacturer though that as of last year has begun supporting Linux from top to bottom with their entire line-up of printers. Not only are they providing CUPS drivers, but also they are even printing Tux in the corner of every box they ship right besides the Windows and Apple logos. Do you know who we are talking about? Probably not, but it's Lexmark. After months of wrangling within the company, Lexmark has stepped up to become a Linux and open-source friendly company. We are seeing how far this Linux support extends as we try out the Lexmark Pro905 Platinum multi-function printer.
According to the release plans, the release of X Server 1.8 should take place, and while in reality it will likely not be released today, its release is coming soon. When this release does arrive, it will add a new set of features to the X.Org stack and a number of other minor improvements and bug-fixes.
Phoromatic, our remote test management system that makes it incredibly simple to deploy the Phoronix Test Suite across an array of systems within an organization or around the world, has been in development for more than a year. We publicly announced this unique enterprise solution when developing Phoronix Test Suite 2.0 and it publicly went into beta with Phoronix Test Suite 2.2 where it became possible to easily build a benchmarking test farm using our Phoronix software. Before ending out the year we launched Phoromatic Tracker with an initial reference implementation to monitor the Linux kernel performance on a daily basis and in a fully automated manner. Phoromatic has been a huge success, but today we are announcing that Phoromatic has reached a 1.0 status and additionally we are providing the Ubuntu Linux community with a new performance tracker in collaboration with Canonical.
The release of Ubuntu 10.04 LTS "Lucid Lynx" is quickly approaching next month and it will arrive with a whole set of new features and improvements including a faster boot process, a long-awaited new theme, the Nouveau driver to replace the crippled xf86-video-nv driver, the unveiling of the Ubuntu One Music Store, integration of Plymouth, Ubuntu ARM advancements, and many other advancements for this Linux distribution. While it may not be as exciting as looking at these new end-user features, in this article we are testing out the available kernels for Ubuntu 10.04. Besides the standard Linux 2.6.32 kernel used in the Lucid release, there is also a specialized server kernel as well as a new -preempt kernel is now available. We are looking at how these different kernels perform and how they compare to the mainline Linux kernels with the 2.6.32, 2.6.33, and 2.6.34-rc1 releases.
Fedora 13 Alpha was released yesterday with a plethora of new features and updated packages for this Red Hat Linux distribution. Aside from the features like Btrfs system rollback support and PolicyKit One support for Qt/KDE applications to excite end-users, each Fedora release always pulls in the very latest Linux graphics code. Fedora was the first distribution shipping with the Nouveau driver, then its KMS driver, and now with Fedora 13 it's the first OS deploying Nouveau's Gallium3D driver (there's benchmarks behind that link). Fedora 13 is also carrying the latest packages for the unreleased X Server 1.8, DisplayPort monitor support for more graphics cards, the latest ATI driver code from the xf86-video-ati DDX to the in-development DRM, and then there is the very latest Intel work too. To get an idea for the direction that the Intel 3D support is heading in this release, we have carried out a few quick OpenGL benchmarks.
Xfce, LXDE, and other desktop environments are often referenced as being lighter-eight Linux desktop environments than KDE and GNOME, but what are the measurable performance differences between them? Curious how much of a quantitative impact the GNOME, KDE, Xfce, and LXDE desktops have on netbook systems, we carried out a small set of tests to look at the differences in memory usage, battery power consumption, and thermal performance.
521 software articles published on Phoronix.