CD recording and RAID

From: Piotr Wajnberg (davine@poczta.onet.pl)
Date: Sun Jan 05 2003 - 11:07:52 EST


I've got a RAID-0 setup and suppose a slowest disk io performance (like 300kb/s) during
CD recording is due to the recorder being on the same ide controller as
my hard disks although md5summing is as fast as usual and can even
starve the recorder causing the recording to fail.
What I'd like to know is why the system completely
freezes for the time a CD is being fixated. Here's the output of
vmstat for the final stages of fixation:
 r b w swpd free buff cache si so bi bo in cs us sy id
 5 0 0 6824 4508 10768 159036 0 0 0 80 172 745 100 0 0
 2 0 0 6824 4508 10768 159036 0 0 0 0 158 741 99 1 0
 7 0 0 6824 4508 10768 159036 0 0 0 0 158 700 100 0 0
 2 0 0 6824 4508 10768 159036 0 0 0 0 158 736 99 1 0
 5 0 0 6824 4484 10768 159036 0 0 0 0 151 703 100 0 0
 3 1 1 6824 4476 10784 159036 0 0 0 12 148 691 100 0 0
 5 1 1 6824 4476 10784 159036 0 0 0 0 151 580 100 0 0
 2 0 1 6824 3416 10808 159036 0 0 0 40 156 639 100 0 0
 6 0 0 6824 8864 10892 154936 0 0 0 253 9384 37130 3 0 97

As you can see from the last entry as soon as the the recording ends the system
catches up, as it were, with all the processes that were cut off from the processor.
I'm currently using 2.4.20-ck2 because I thought the patches would cure the
condition but they've made no difference. Could anyone, please, offer an explanation?

-- 
PeteVine
  5:07pm  up 22:54,  2 users,  load average: 1.11, 1.18, 1.27
65 processes: 60 sleeping, 5 running, 0 zombie, 0 stopped
CPU states:  3.9% user,  5.8% system, 90.1% nice,  0.0% idle

Linux registered user #250250 http://counter.li.org

--------------r-e-k-l-a-m-a-----------------

Super tanie kwatery narciarskie. Od 300 zl/osoba/tydzien http://wycieczki.onet.pl - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Tue Jan 07 2003 - 22:00:28 EST