Home
Reading
Searching
Subscribe
Sponsors
Statistics
Posting
Contact
Spam
Lists
Links
About
Hosting
Filtering
Features Download
Marketing
Archives
FAQ
Blog
 
Gmane
From: Daniel Poelzleithner <poelzi <at> poelzi.org>
Subject: [ANN] ulatencyd 0.3.1
Newsgroups: gmane.linux.kernel
Date: Tuesday 18th January 2011 10:57:08 UTC (over 5 years ago)
Hi all,

I'm happy to announce the release of ulatencyd 0.3.1

What is it ?
============

ulatencyd is a scriptable daemon which constantly optimises the Linux
kernel for best user experience.
The default configuration tries reduce the latency for a typical desktop
system and protects the system from malicious processes/groups.
With a different configuration all other types of systems can be
adjusted as well.

Website: <https://github.com/poelzi/ulatencyd/>

• Ubuntu PPA: <https://launchpad.net/~poelzi/+archive/ulatencyd-stable>
• Latest Release:
<https://github.com/downloads/poelzi/ulatencyd/ulatencyd-0.3.1.tar.gz>


What's new in 0.3 (very short version)?
=======================================

• X-Server plugin which monitors the active windows and gives more
  priority to the recently active windows.
• Full scripted scheduler with flexible configuration that allows easy
  adjustment of the resulting cgroups and parameters.
• Protection rules against common cases of memory shortage, also known
  as the swap of death (even without swap).
• Scheduler configuration for desktop systems
• Fast C core with Lua embedding. Intelligent caching to reduce the
  load on filter rules and scheduler.
• Very detailed API for system inspection and adjustment.
• Rudimentary Rules for popular Desktop Environments KDE & Gnome


Does it work ?
==============

Yes. Of course it is possible to still cause a swap of death or overload
the system so much, that it starts to get lag. But typical cases get
caught by the current rules. A make -j 50 of the linux kernel
on my dual core machine with 4 GB ram let me still work on the machine
quite well, but a make -j 80 still locks it down for like 4 minutes.
This is most likely because the heavy swapping done on the isolation
groups (their main memory usage is restraint). This may be fixed by
using the blockio subsystem, but this is not used yet. Planned for the
next version.


kind regards
 Daniel Poelzleithner
 
CD: 3ms