FWN/Issue200

= Fedora Weekly News Issue 200 =

Welcome to Fedora Weekly News Issue 200 for the week ending November 1, 2009. What follows are some highlights from this issue.

Welcome to FWN issue 200, an impressive milestone! This week's issue starts off with news and views from the Fedora community, including further work on libguestfs, examination of several new features in Fedora 12, and work on a new tool for ICC color management in Gnome. In Quality Assurance, details from last week's Test Day on internationalization support in Fedora, and great updates on the various QA weekly meetings as we get closer to Fedora 12. In Translation news, several updates pertinent to Fedora 12 GA release, as well as details on Publican 1.0, which the Docs and Transaltion teams use for publishing books, articles, papers and multi-volume sets with DocBook XML. In Design news, details on the final Fedora 12 wallpapers, decisioning on extra wallpapers for the release, and some thoughts on the F12 art process looking forward to the next cycle. Security Advisories brings us current on the numerous security patches released this past week for Fedora 10 and 11. Our issue wraps up with news from the Fedora virtualization and libvirt lists, including a recent summary of Fedora virt bugs and developments, the state of KSM tuning on Fedora systems and a couple items on QEMU related issues with monitor handling and QEMU driver thread safety rules. Please enjoy Fedora Weekly News issue 200!

If you are interested in contributing to Fedora Weekly News, please see our 'join' page. We welcome reader feedback: fedora-news-list@redhat.com

The Fedora News team is collaborating with Marketing and Docs to come up with a new exciting platform for disseminating news and views on Fedora, called Fedora Insight. We plan to have the next issue of Fedora Weekly News in Fedora Insight, next week. We welcome your feedback as we migrate FWN to this new content platform!

FWN Editorial Team: Pascal Calarco, Adam Williamson

>= 1.0.75 (from updates or koji )
 * a few minutes of your time

service must be enabled: sudo chkconfig ksm on

Mark McLoughlin also noted The maximum number of pages which may be merged defaults to half of the system memory, and may also be manually defined in. "Here's the logic we have in the init script :" # unless KSM_MAX_KERNEL_PAGES is set, let ksm munch up to half of total memory. default_max_kernel_pages { local total pagesize total=`awk '/^MemTotal:/ {print $2}' /proc/meminfo` pagesize=`getconf PAGESIZE` echo $[total * 1024 / pagesize / 2] }

Justin Forbes points out "The limit to half of total memory is because ksm pages are unswappable at this time. To be fixed in a future kernel."

A second service,, may also be enabled. Ksmtuned regulates how aggressively the system will attempt to merge pages. Parameters such as how many pages to scan before sleeping and how long to sleep may be configured in.

Memory pages must be flagged as mergable before KSM will scan them looking for duplicates. At present only Qemu pages will be marked as such. As described in the kernel docs , the effect of KSM system memory may be examined in. "A high ratio of pages_sharing to pages_shared indicates good sharing, but a high ratio of pages_unshared to pages_sharing indicates wasted effort."

Libvirt List
This section contains the discussion happening on the libvir-list.

Node device enumeration with udev
Dave Allan posted "a fully functional version of the node device udev based backend, incorporating all the feedback from earlier revisions." "...I have also included a patch removing the DevKit backend."

Also see FWN#146 "Host Device Enumeration API" for some coverage of the host device enumeration API.

Rewrite of QEMU monitor handling
Daniel Berrange posted a "patch series [which] rewrites the QEMU monitor handling almost completely.

The key theme here is to move from a totally synchronous way of interacting with the monitor, to a totally asynchronous way. This allows " " to handle receipt & dispatch of asychronous events from QEMU. For example a notification of a disk-full error, or VM state change. In the process of doing this re-factoring I have also dropped in basic support/infrastructure for the JSON based monitor."

Libvirt QEMU driver thread safety rules
In a characteristically long and detailed post Daniel Berrange laid down the law on thread safety rules for the Qemu driver.

"This document describes how thread safety is ensured throughout the QEMU driver. The criteria for this model are:


 * Objects must never be exclusively locked for any pro-longed time
 * Code which sleeps must be able to time out after suitable period
 * Must be safe against dispatch asynchronous events from monitor"

Also see FWN#155 "Thread Safety for libvirtd Daemon and Drivers"