I currently only use it on 8 machines, 3 linux the rest win2k
the webpage, it isnt as spectacular as other operating systems
but, all looks well to me other than the 50% thing in the previous email
they are on the same multi-cast they seem to co-exist fine group.
most of the code was from machines/linux.c and i added a few cygwinism's
and left stubs for things cygwin /proc doesn't provide
well, it'd probably break if metric_typedefs.h changes,
I doubt that any patch to linux.c would apply cleanly without something
like a cleaver whitespace trickery to linux.c and cygwin.c to make
everything on the same lines.... still then hoping that the patches
aren't in the gray area where cygwin.c and linux.c differ
from a users point of view, cygwin doesn't yet have all the components
required to build it..
I hope to alleviate this though.
I did a diff, between linux.c and cvs linux.c It didn't seem like any of
the modifications made since 2.4.1 would
be terribly hard to backport to the cygwin.c still though, it's not...
automatic
I see 2 ways of going about maintanance
maintaining 2 seprate files ala cygwin.c, or fixing cygwins /proc/
add /proc/cpuinfo
and add the 1/999 9999 code to/proc/loadavg /proc/loadavg
and add cpu0,cpu1 information code to /proc/stat
and then symlinking machine.c to linux.c... sounds easier to maintain
and harder to implement...
I'll work on getting sunrpc into the cygwin dist.. I've already improved
the rpclib patch quite a bit.
it's almost ready to go up for a vote on the appropriate list.
matt massie wrote:
matt-
i haven't replied to this because i don't really know what to say. i'm
really impressed that you got ganglia running on windows! i'm shocked
even. i'd like to fold in windows support as part of the standard
distribution if possible.
how hard would it be to maintain? how much have you tested the code and
used it? how stable is it? how many machines have you run it on?
when i tried to port ganglia to cygwin (granted i only spent a day trying)
i found that cygwin didn't support pthreads so i didn't get all the
locking, threads, condition variables etc that ganglia uses throughout the
code.
great work! i'll take a look at this as soon as i can.
-matt
Tuesday, matt rice wrote forth saying...
notes:
configure.in.patch... don't apply this under cygwin... unless you
feel like fixing all the acconfig.h things.. Makefile will notice
configure.in is not up to date and try to re-generate things, which
won't work. I add -lrpclib to LIBS in gmond/Makefile and
gmetric/Makefile (also this means I haven't tried it)
rpcinfo.c.patch: apply this to
http://grass.itc.it/grass5/binary/windows_cygnus/sunrpc/sunrpc-4.0.cygwin1.src.tar.gz
cygwin.c.patch:
there are bugs, a seg fault when running in gdb,
cpu_system seems to sit at 50% if nothing is going on
it won't run without the attached /etc/gmond.conf..
tip:
cygrunsrv -I gmond -f "Ganglia Monitoring Daemon" --path
/usr/sbin/gmond.exe
will add gmond as a NT Service..
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers
