vital signs in hijack

Posted by: image

vital signs in hijack - 23/06/2002 08:30

i recently installed emptriv. i wanted to see how much processor power it was taking to unzip the whole thing. so while it was happening, i put on vial signs. says it was only taking up 1.58% of cpu. can this be? or does it only show the player's processor power?
Posted by: Shonky

Re: vital signs in hijack - 23/06/2002 19:13

These are not CPU usage numbers. They are "load averages". The load averages are the average number of process ready to run during the last 1, 5 and 15 minutes.
Posted by: mlord

Re: vital signs in hijack - 28/06/2002 12:15

Load numbers larger than 1 usually indicate that the CPU is VERY BUSY.

-ml
Posted by: smu

Re: vital signs in hijack - 30/06/2002 04:23

Hi Mark.

I just wondered about a few things while installing a new (software) raid-5 IDE array in my Linux box. I thought you might be able to comment on these:
  1. While copying files to the raid (and having an rc5 client run in the background at maximum nice level), I noticed that the load went up to 5 or even higher, but with CPU usage of approx. 60% system, 20% user and 20%idle. While I can understand the first two, I have no idea why I could have a load of 5 with 20% idle time. Any hint?
  2. I also noticed that the raid driver reports a few stats in /proc/mdstats, the particular thing that made me look there twice was that it reported on its writing speed during resynchronization of the raid discs. Don't remember the exact phrase used, but it showed a nice little progress bar, along with a textual comment like "X% done, Y minutes remaining, writing in background with ZZZZZ kBytes/second". It was writing at about 22MBytes/s on an idle system, and about 15 on a loaded one. Nice. I was wondering if the IDE driver could report its current throughput in a similar manner.
Any comments from you?

cu,
sven
Posted by: mlord

Re: vital signs in hijack - 01/07/2002 07:01

While I do not recall the EXACT calculation for "load average", my recollection is that it is the average number of processes waiting for either CPU or IO (usually disk). So it is very common to see a high load average with low CPU utilization.

The RAID statistics in /proc/ are there because the RAID code is deliberately self-throttling while rebuilding. The statistics serve as a tuning indication only, in case one wants to slow it down even further.

Nearly all block I/O drivers (including IDE) keep similar numbers in the kernel, and there may exist userland applications for displaying them. The IDE driver doesn't have the code bloat that would be needed to calculate & display them in /proc/ (but these days it does have lots of other bloat..).

Cheers