[lkml]   [1996]   [Apr]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: 1.3.86 - Stability under high loads: *GOOD*
On Thu, 11 Apr 1996, Nuno Serrenho wrote:

> I type 'uptime' and find these wonderful values:
> 12:24pm up 14:02, 2 users, load average: 26.43, 37.54, 38.97
> I've since cleaned-up the script so that this *won't* happen again, but it
> sure shows that stability is quite good under high loads. Most other
> OS's (and I'm not going to mention any names here!) would probably grind to
> a halt with sysloads of 30+ for 5 hours straight.

i just recently put my system through a real torture test to check the

i ran the following all concurrently:

2 "find / -name \*.c -exec /usr/bin/grep . '{}' > /dev/null';'"
104 md5's on a 30 meg file
compiled gcc (stage 3 compiler for c, c++, obj-c & fortran)
tarred & gzipped an 80 meg source directory

and the system was usable the whole hour it kept a load average over 50.
later that night i ran 1000 concurrent md5's while building gcc (same
config) and the next morning i had to kill off the remaining md5's so i
could use the system. but it did recover from a load average just over
255 (for about 4 hours) and i didn't have any processes die or anything.

in both cases, all my md5 processes matched, my gcc compile (when it
completed) was fine according to "make compare", & the directory i
tarred & gzipped was fine. oh, and most importantly, the system was
STILL stable.

simply put, linux is one HELL of an operating system.

PS: the above tests were on linux v1.3.83, IDE system with ppp up.
Why Linux? source code. POSIX. tcpip. job control. support from the authors.
drivers for most hardware. because one terminal or process is never enough.
forget the other O/Ss, i use Linux- the choice of a GNU generation.

 \ /
  Last update: 2005-03-22 13:37    [W:0.053 / U:2.964 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site