October 25, 2014
 
 
RSSRSS feed

Linux 3.10 Improves Multi-tasking and SSD Caching

  • July 1, 2013
  • By Sean Michael Kerner

The third major milestone release for the Linux kernel in 2013 is now available. Linux 3.10 The new release follows the previous Linux 3.9 kernel release by just over two months.

"In the bigger picture (ie since 3.9) this release has been pretty typical and not particularly prone to problems, despite my waffling about the exact release date," Linux creator, Linus Torvalds wrote in release message. "As usual, the bulk patch-wise is all drivers (pretty much exactly two thirds), while the rest is evenly split between arch updates and "misc". No major new subsystems this time around, although there are individual new features."

SSD Caching

As was the case with the Linux 3.9 kernel, SSD's once again are front and center. For the Linux 3.10 kernel Google Linux kernel developer, Kent Overstreet landed the bcache, block layer cache technology for Linux.

"Does writethrough and writeback caching, handles unclean shutdown, and has a bunch of other nifty features motivated by real world usage," Overstreet wrote in his commit message.

Multi-tasking

Improving multi-tasking capabilities in Linux has been a key focus for many years. Back in 2007 the Linux 2.6.23 kernel for example implemented the Completely Fair Scheduler (CFS).

In Linux 3.10 we now get what is known as, 'Timer Free Multitasking'.

"This feature extends the nohz variable-size timer tick feature from idle to busy CPUs (running at most one task) as well, potentially reducing the number of timer interrupts significantly," Linus Torvalds wrote in his commit message.

The net effect of the timer is improved performance. According to Torvalds, with this new capability performance could be improved by up to 1 percent for certain workloads.

TCP Optimization

Linux is widely used by networking vendors and as such the TCP stack within Linux has been a key focus for improvement as well. In Linux 3.10, the IETF Tail Loss Probe is being implemented in Linux.

"Retransmission timeouts are detrimental to application latency,especially for short transfers such as Web transactions where timeouts can often take longer than all of the rest of a transaction," IETF's draft spec for Tail Loss Probe explains. " The primary cause of retransmission timeouts are lost segments at the tail of transactions."

Google Linux kernel developer Nandita Dukkipati explained his commit message that, TLP's goal is to reduce tail latency of short transactions. It achieves this by converting retransmission timeouts (RTOs) occuring due to tail losses (losses at end of transactions) into fast recovery.

Sean Michael Kerner is a senior editor at Linux Planet and InternetNews.com. Follow him on Twitter @TechJournalist.

Sitemap | Contact Us