Re: high average load and kernel recompilation

On Sat, 2009-02-21 at 18:58 +0100, Pol wrote:
It is hp tc4200 laptop, 1.7 GHz

Ah, a tablet. I know that one from the company, it's pretty slow. It's a
single-core CPU, right (Intel Pentium M 760?)

(but i am used to keep it in 'powersave'
mode, at 0.8GHz) ,

What does this mean? AFAIK Ubuntu (and I would guess Kubuntu) uses an
on-demand CPU governor by default. That is, it will let the CPU run at
full speed whenever needed, but throttle it otherwise. This is
independent from whether the laptop is plugged in or not.

So, my question would be, did you change that, and do you keep your CPU
throttled to a fixed speed even if you have high load? In this case you
should look no further for now and see whether the load continues to be
so high on average even it the CPU runs at full speed. After all, you'd
be throttling it to less than half its (less than spectacular as it is)
speed.

As i log in, several system services are started: postfix, dovecot, dictd,
hpoj, wooffle, collectd. I one of my attempts, proxy was disabled in
webbrowsers, with no change in performance, so my guess is that wwwoffle
should not affect the average load.
After log in, a few applets (kbandwidth, clock, calendar, systray), two
konsoles monitoring log files with 'tail -f', amarok are started and, from
a few more konsole sessions, 'sudo su' to four more users.
So, there are five users, each of which starts konqueror, connects to gmail,
opens a few pdf files, starts kalarm and kgmailnotifier; sometimes lyx
editors and maxima (symbolic computations, spending most of the time as
idle) are started.
One of those users starts ktorrent; another one starts knode; another one
starts firefox (connecting to gmail and opening several google notebooks)

Um. You are running 5 users concurrently from a tc 4200 (possibly even
throttled?) and you complain about an average load of 5?

I run one user on an HP nc6400 (2GB RAM) with evolution, terminal(s),
firefox, liferea, video player, and transmission. I currently have load
average: 0.22, 0.81, 1.42

I dunno, to me this sounds like "to be expected", but maybe someone else can chip in,

2.6.24-23-generic

That's fine, you won't get any noticabe difference over the "generic"
flavor by recompiling, I don't think.


--
ubuntu-users mailing list
ubuntu-users@xxxxxxxxxxxxxxxx
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

Relevant Pages

  • Re: Fwd: Re: New Metrics to measure Load average
    ... School of Electrical and Information Engineering ... Currently for the Phd, I am using division of load as per login users and this is what I want to have for HPC and for the "Task Profiling Model". ... Be very wary if their planned CPU controller as the model that they intend to use will only work with CPU bound processes. ...
    (Linux-Kernel)
  • [PATCH 2/2] sched: Improve fairness of cpu allocation for task groups
    ... The current load balancing scheme isn't good for group fairness. ... Col4 -> CPU bandwidth received by each individual task of the group. ... Currently cpu's weight is just the summation of individual task weights. ...
    (Linux-Kernel)
  • [PATCH] sched: Consolidated and improved smpnice patch
    ... The introduction of separate run queues per CPU has brought with it "nice" enforcement problems that are best described by a simple example. ... whether this expectation is met is pretty much down to luck as there are four equally likely distributions of the tasks to the CPUs that the load balancing code will consider to be balanced with loads of 2.0 for each CPU. ... To overcome this problem SCHED_LOAD_SCALE has been replaced by the load_weight for the relevant task or by the average load_weight per task for the queue in question. ...
    (Linux-Kernel)
  • [PATCH 2/2] sched: Improve fairness of cpu allocation for task groups
    ... The current load balancing scheme isn't good for group fairness. ... Col4 -> CPU bandwidth received by each individual task of the group. ... Currently cpu's weight is just the summation of individual task weights. ...
    (Linux-Kernel)
  • Re: New Metrics to measure Load average
    ... Our project which is conducted as part of CELrequires to separate the load along individual user lines. ... We have developed a prediction model called "Free Load Profile Model for Load Profile Prediction" for the purpose of scheduling in the Grid. ... The problem with the load metric of current Linux/Unix is that it measures CPU load and Disk load without indicating the true nature of the load, thereby creating some confusion among the readers. ...
    (Linux-Kernel)