CACTID Issues - Windows [RESOLVED]

Post by **TheWitness** » Mon Jun 04, 2007 7:17 am

It would appear that Cygwin still can't get rid of those bugs in their command interpretor. Shame from my perspective as well.

One more question, did each sh.exe have a corresponding cmd.exe as well?

Larry

mcutting · Post by **mcutting** » Mon Jun 04, 2007 7:19 am

Hi Larry,

Yes, as it happens, there were CMD.exe process along with SH.exe. I am running the G version of the poller, and all is well at less than 26 seconds for 47 hosts !

Post by **TheWitness** » Mon Jun 04, 2007 7:36 am

Do you get any warning messages in the log around the time these processes lock up?

Larry

mcutting · Post by **mcutting** » Mon Jun 04, 2007 7:44 am

No - nothing at all.

Post by **TheWitness** » Mon Jun 04, 2007 7:55 am

I'm looking at Cygwin source code right now... I hate Windows by the way, and I've been using it since '84, but I'll do this now.

Larry

mcutting · Post by **mcutting** » Mon Jun 04, 2007 7:56 am

Great - thanks

Post by **TheWitness** » Mon Jun 04, 2007 8:48 am

mcutting,

In the following poller.c, I am using the native Cygwin popen command. Can you recompile and see if this provides any joy.

Larry

mcutting · Post by **mcutting** » Mon Jun 04, 2007 9:27 am

Hi Larry,

I get the following error when I replace the poller.c file with the one you have posted

poller.c: In function `exec_poll':
poller.c:924: error: `MAX_SIMULTANEOUS_SCRIPTS' undeclared (first use in this function)
poller.c:924: error: (Each undeclared identifier is reported only once
poller.c:924: error: for each function it appears in.)
make: *** [poller.o] Error 1

I also can't make install...

For info, I decided to compile a copy of CACTID 0.8.6i which is availble from www.cacti.net/downloads to see if this makes any difference..

Any ideas ?

Thanks

Post by **TheWitness** » Mon Jun 04, 2007 12:56 pm

It would appear as if you are not running the SVN version of the code. Let me attach it now. Or, maybe I have not committed it

Who knows...

Larry

mcutting · Post by **mcutting** » Tue Jun 05, 2007 12:35 am

Hi Larry,

Managed to compile this one, however, cactid.exe exhibits the same problem. It actually failed on the first run, with sh.exe taking 99 percent cpu, and crashing the shell.

It seems that the G version works fine, as it doesn't use sh.exe. I have rolled back to this version, as it works flawlessly.

However, I'd still like to use the most up to date version - any ideas ?

Thanks

Post by **TheWitness** » Tue Jun 05, 2007 8:11 am

Well, this would appear to be a bug in either windows or Cygwin. Don't know which. It would appear as if some scripts either don't shut down cleanly, or there is a interprocess communications problem.

Without seeing it myself, I will be hard to troubleshoot. You would have to provide access in order for me to investigate further.

Larry

mcutting · Post by **mcutting** » Tue Jun 05, 2007 8:30 am

Ok. One question - why was sh.exe reintroduced in versions above G (cactid) ? The G version of the poller works fine ?

Post by **TheWitness** » Tue Jun 05, 2007 11:09 am

Supprisingly, to reduce problems in Windows and specific environment related issues in other OS'. Since Winows is so unstable (Namely Cygwin), I may have to put it back in.

Larry

mcutting · Post by **mcutting** » Tue Jun 05, 2007 11:29 am

Cool - can I be one of the testers if you are going to put it back in ?

Thanks Larry.

mcutting · Post by **mcutting** » Mon Jun 11, 2007 3:12 am

mcutting wrote:Cool - can I be one of the testers if you are going to put it back in ?

Thanks Larry.

Hi Larry - any possibility that an "i" version of the poller could be released that has sh.exe rolled in (as "g" does, which works fine) ?

Thanks

Cacti

CACTID Issues - Windows [RESOLVED]

Who is online