[prev in list] [next in list] [prev in thread] [next in thread] 

List:       grid-engine-users
Subject:    loadcheck
From:       "Tim Harsch" <harsch1 () llnl ! gov>
Date:       2003-01-22 21:23:52
Message-ID: 00a001c2c25c$8f9a2960$88bb0986 () BBRP2000 ! LLNL ! GOV
[Download RAW message or body]

My platform = 5.3p1

I was looking into load balancing a bit.  Something I worked on awhile back
but had to back burner.  I notice that in $SGE_ROOT/utilbin/<arch> there is
a 'loadcheck' program.  I assume this intended to be a default load checker.
But
in reading the man on sge_execd I found that a load sensor has a required
output format.  Here is an example run of loadcheck:

[49] trna:/cdna3/apps/SGE/utilbin/solaris64 {root/harsch}# loadcheck
arch            solaris64
num_proc        8
load_short      1.09
load_medium     1.02
load_long       0.75
mem_free        3277.000000M
swap_free       37066.000000M
virtual_free    40343.000000M
mem_total       32768.000000M
swap_total      40539.000000M
virtual_total   73307.000000M
mem_used        29491.000000M
swap_used       3473.000000M
virtual_used    32964.000000M
cpu             13.0%

Here is the documentation from man sge_execd that describes the output
format for a load sensor.  Notice that they do not match.

     o  A load value report starts with a line containing nothing
        but the word start.

     o  Individual load values are separated by newlines.

     o  Each load value report consists of three parts  separated
        by colons (":") and containing no blanks.

     o  The first part of a load value information is either  the
        name  of  the host for which load is reported or the spe-
        cial name "global".

     o  The second part is the symbolic name of the load value as
        defined  in  the  host  or  global complex list (see com-
        plex(5) for details). If a load  value  is  reported  for
        which no entry in the host or global complex list exists,
        the reported load value is not used.

     o  The third part is the measured load value.

     o  A load value report ends with a line with the word end.


What am I missing?

Tim Harsch
Computer Scientist
Lawrence Livermore National Laboratory
(925) 423-1135


[prev in list] [next in list] [prev in thread] [next in thread]