In association with heise online

File systems

Some of the optimisations of VFS (Virtual File System), which the userspace uses to access file systems, in 2.6.38 were especially important for Torvalds, who could not hide his excitement in a detailed description in an email on the first pre-release version of 2.6.38 (see 1, 2, 3, 4, 5, 6, 7). Due to these the VFS now uses finer locking with RCU (read copy update). Thanks to changes developed under such names as "RCU-based name lookup", a number of operations in the resolution of file names are considerably accelerated. Large servers with a number of processor cores will not be the only ones to benefit; off-the-shelf systems will too. In his release email, Torvalds says the performance improvements range from 30 to 50 per cent in certain tests of file and name resolution. In an earlier email on the adoption of deep optimisations of VFS, which is very important for the reliability of the kernel, he wrote that a Find command executed with filled caches in his home directory was around 35 per cent faster, even though Find was only working with a single thread.

In 2.6.38, Ext3 and XFS now support batched discard which was integrated in the 37 kernel and is especially interesting for SSDs with a slow TRIM function (1, 2, 3). The still experimental CoW (Copy on Write) file system Btrfs can use LZO, in addition to Zlib, for transparent compression. It is generally much faster, but does not compress quite as efficiently; the commit comment provides some of the measurements supporting that claim along with a comparison with an uncompressed file system (1, 2). Btrfs now also supports write-protected snapshots ; shortly before the release of RC7 the kernel developers fixed a flaw in the Btrfs code; it had been discussed earlier on the LKML and could cause recently created files to be empty when copied with current versions of the "cp" coreutils component.

Used as a file system in a number of Linux distributions for live and installation media, SquashFS in 2.6.38 supports the compression of file system images with the XZ compression format, spun out of LZMA (1, 2). Because it is known for very great compression density, some live media will probably be able to offer a few more programs than before. Used with optical media (such as DVDs), the UDF file system does without the Big Kernel Lock (BKL) starting in 2.6.38. The developers have therefore removed BKL from the last widely used kernel component that still required BKL, following the BKL removal campaign in 2.6.37. There is also a rough roadmap to rid the world of the remaining users of BKL (1, 2).

Downloading the Linux kernel

New versions of Linux can be obtained from the Kernel.org servers; the contents of these servers are also mirrored on numerous mirrors internationally. However, Linux users who are not familiar with the details of the kernel and its environment should generally not install new Linux drivers and kernels themselves but use the kernels provided by their Linux distributors instead.

Storage

The kernel developers have added the LIO multi-protocol storage target developed by linux-iscsi.org to the SCSI subsystem. This allows storage devices to be integrated as SCSI targets which other systems can access via iSCSI, Fibre Channel or FCoE (Fibre Channel over Ethernet). This solution for implementing SANs (Storage Area Networks) will soon replace STGT (SCSI Target Framework; kernel configuration option CONFIG_SCSI_TGT), which has been part of the kernel for some time. The developers of SCST (SCSI Target Subsystem for Linux) had also promoted their solution as a replacement, but were unsuccessful. A short overview of LIO's features can be found in the commit comment. Some background information on the technical differences between the various SCSI target implementations and on the aspects responsible for the decision in favour of LIO can be found in the article "A tale of two SCSI targets" on LWN.net.

The use of multiple work queues in Dmcrypt is designed to improve the Device Mapper's (DM) handling of multi-core processors and should speed up encryption and decryption. New interfaces enable the Device Mapper to access the MD (Multiple Devices) subsystem's software RAID admin functions for RAID levels 4, 5 and 6. Together with a fresh version of DMRAID this should, in the long term, improve the kernel's support for the host RAID 5 features offered by many motherboard chip-sets. These features were previously only accessible via special kernel patches that aren't included in all distributions.

Next: Networking and infrastructure

Print Version | Permalink: http://h-online.com/-1205467
  • Twitter
  • Facebook
  • submit to slashdot
  • StumbleUpon
  • submit to reddit
 


  • July's Community Calendar





The H Open

The H Security

The H Developer

The H Internet Toolkit