[Date Prev]   [Date Next] [Thread Prev]   [Thread Next] [Date Index]   [Thread Index]


     Re: [nocol-users] Strange webnocol behavior

>On Fri, Jul 14, 2000 at 01:59:17PM -0400, JoJo Brooks wrote:
>> Has anyone else here seen instances where genweb.pl will sometimes
>> fail to completely generate the Info.html page?
>> I haven't yet been able to pin down a common cause, but it will
>> happen at different places in the readout, lately 2 or 3 times
>> an hour (being updated every 5 minutes).
>> I haven't seen any difference in the appearance of the same information
>> using netconsole, so I'm starting to wonder if the problem may be
>> being caused by something coming back from the remote hostmon-client,
>> or even someplace else...
>How big is your info page?  What sort of system are you running on?
>What version of NOCOL?  What version of Perl?  Which browser are you
>using?  Have you compared the file against what you see in the browser
>(ie. the HTML source)?  etc, etc...

The Info page used to be about 160-165 info lines.  Earlier it was only
truncating every now and then, but now it craps out each and every update,
done automatically or manually, all while generating the hostmon info.

The line of death appears now to be consistent, around line 60-61 each 
time now, even if I swap the order of machines around.  If I remove 
hostmon, all of the other monitored information displays correctly.

The last few characters in the HTML file seem to vary, but it appears 
to always die around the same place, while writing either the 
"command=Updates" or "return=http://" part of the line, so yes, the HTML 
files are indeed truncated, meaning that it's not my browser that's funky 
(no javascript used either!).  And all the info does appear correctly in
netconsole, so the information coming back from the various servers is 
not being truncated (which is what I first thought the problem was).

Currently using v4.3.1, and I literally haven't had to touch it since I 
first installed it. Also using perl 5.004_04, and Netscape 4.08, Solaris 7
on a Sparc10 (sun4m), and all had been running find since Day 1 of this 
install, until this afternoon.

I cannot find any other differences, and definitely nothing in any of
the perl scripts that generate the pages has changed.

Someone did make use of the Updates feature while I was on vacation
because they thought a Critical notice wasn't being updated fast enough
(meaning they fixed the problem, but the update back to Info didn't happen 
fast enough for them), but as far as I can tell, the choking only started
today, nearly two weeks after that incident.

I'm at a loss at what else to check.

JoJo Brooks                                     
Advanced Systems Consulting, Inc                    jbrooks at advsys dot com