Alex:
Thanks! The crash problem can be found in 5.2 and 5.0.9 too. I have feeling the
5.0.9 may be still okay because I manually tar the /usr/local to the target machine.
And there are some libxx.5.1.1 and soft link staff in /usr/local/lib. I suspect my
installation of 5.0.9 has mixed up with the left over 5.1.1 lib files in the
/usr/local path. I am not 100% sure the crash indeed happens on 5.0.9. Since I tested
5.0.6 before and it's save. And I have the INSTALL_PREFIX option added in 5.0.6. So
the tar ball should be good for installation. I will just take 5.0.6 for now. Thanks
a lot for the direction!
-Jim
-----Original Message-----
From: Alex Burger [mailto:[EMAIL PROTECTED]
Sent: Monday, September 27, 2004 8:39 PM
To: Jim Su
Cc: [EMAIL PROTECTED]
Subject: Re: win32 5.2.pre1 snmpwalk times out
Hi Jim.
You can download the current version from the CVS repository.
Instructions can be found here:
http://sourceforge.net/docman/display_doc.php?docid=2772&group_id=12694
Alex
Jim Su wrote:
> Hi Alex:
>
> AB> I installed the cvs version of 5.2 from 9/22/04.
>
> I found the net-snmp-5.2.pre1.tar.gz from the web and the date is 9/2/2004. Where I
> can get the 5.2 on 9/22/2004?
>
> Thanks,
> Jim
>
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Alex Burger
> Sent: Saturday, September 25, 2004 7:41 AM
> To: [EMAIL PROTECTED]
> Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Re: win32 5.2.pre1 snmpwalk times out
>
>
>
> Hi.
>
> While setting up a new machine with Windows 2000 Professional, I decided
> to do some testing.
>
> I installed the cvs version of 5.2 from 9/22/04. The following worked
> when I did a full walk of .1:
>
> -fresh W2k Pro with SP1
> -after installing SP4
> -after running Windows update, scanned for updates and let it install IE6
>
> So up to that point, it seemed to work fine.
>
> After I installed the following, the problem I reported started:
>
> -ran Windows update, scanned for updates and let it install IE6
> cummulative update plus a bunch of hotfixes (critical updates & service
> packs). As listed in the event log:
>
> Q828026
> KB840315
> KB842526
> KB841873
> KB841872
> KB839643
> KB839645
> KB837001
> KB828741
> KB835732
> KB329115
> KB828035
> KB825119
> KB826232
> KB824105
> KB823182
> KB823559
>
> I do not know which patch broke it, but at least I can re-produce it. I
> will have to do some more testing.
>
> Before this last test, two out of four machines were fully patched based
> on Windows update. The other two only had a select list of critical
> updates applied including the latest IE6 patch.
>
> Looking through the descriptions of the hotfixes, nothing obvious jumps
> out at me as to what is causing the problem.
>
> Alex
>
>
>
> [EMAIL PROTECTED] wrote:
>
>
>>Alex, I think the problem could use some illumination.
>>########################
>>Hi, net-snmp-users!
>>
>>Some of the developers are busy preparing a 5.2.pre2 release.
>>
>>One area that is getting attention is registering the
>>Agent and Trap Handler as Windows Services. This will help
>>using these programs a little easier in the Win32 environment.
>>
>>During testing, Alex noticed that the 5.2.pre1+ Agent would use
>>an inordinate amount of CPU time during "walk" through the
>>network related MIB objects.
>>
>>Alex notes the problem did not occur with the 5.1.2 Agent.
>>
>>I'm just posting this to the mail list to archive
>>the discussion.
>>########################
>>
>>Alex, I would be happy to be proven wrong (in this case!), but
>>it seems that the problem is not the net-snmp agent,
>>rather an interaction between Microsoft IP Helper library
>>and the network adapters. I believe if you were to
>>debug the net-snmp agent, you will find the "hang" somewhere
>>in a call to the IP Helper Library.
>>
>>Have you tried a debugging build to produce symbol map, etc. ?
>>
>>Best Regards,
>>-Mike
>>
>>
>>
>>
>>>From: Alex Burger <[EMAIL PROTECTED]>
>>>Date: 2004/09/22 Wed PM 10:41:17 EDT
>>>To: [EMAIL PROTECTED]
>>>CC: [EMAIL PROTECTED]
>>>Subject: Re: patches are applied
>>>
>>>
>>>Hi again.
>>>
>>>I just tried on yet another machine: my Windows 2000 Server that is
>>>running inside a VMWare on my Linux box (AMD Athalon 1.4Ghz). I get the
>>>exact same problem.
>>>
>>>I have always installed to c or d:\program files\net-snmp. I installed
>>>to c:\usr this time to be safe, and used public as a rw community string
>>>and the same problem.. Same thing with -v 1 and -v 2.
>>>
>>>The spot that it dies from what I can see is the entries in the ARP
>>>table. I just did another test on my main machine and found if I
>>>disable my LAN adapter, it dies at this location instead:
>>>
>>>IF-MIB::ifOutDiscards.1 = Counter32: 0
>>>IF-MIB::ifOutErrors.1 = Counter32: 0
>>>IF-MIB::ifOutQLen.1 = Gauge32: 0
>>>IF-MIB::ifSpecific.1 = OID: SNMPv2-SMI::zeroDotZero
>>>Timeout: No Response from localhost
>>>
>>>That was from doing a walk on localhost, which is always available.
>>>Maybe you could try that. 5.1.2 works fine. It returns nothing (which
>>>is correct), but it doesn't timeout or spike to 99%.
>>>
>>>One thing in common between my two work machines I tested with is that
>>>they both had the Cisco VPN client installed. It's not running, but
>>>there is a LAN adapter for the VPN client.
>>>
>>>Something in common between my home machine and my VMWare Windows is
>>>that they have both had IP addresses assigned to a network card that was
>>>removed without being uninstalled. A new network card has been
>>>installed and the same IP address assigned. Inside the registry it has
>>>both adapters still, and when I change the IP info it always says that I
>>>already got one configured with that IP address, even though I really
>>>don't. You can't uninstall a network card that does not exist, so I'm
>>>stuck with it.
>>>
>>>It probably has nothing to do with it, but who knows. I mentioned
>>>similarities between my two home and my two work computers, but there is
>>>not anything in common between all four... I don't have this problem
>>>with 5.1.2.. I just tested 5.1.2 in my VMWare Windows and it worked fine.
>>>
>>>You should try multiple times. One time during testing it seemed to
>>>work, and then the next time it didn't. It fails 100% of the time on my
>>>now, but before I did have a couple successful ones.
>>>
>>>Alex
>>>
>>>
>>>[EMAIL PROTECTED] wrote:
>>>
>>>
>>>
>>>>I'm using your binary. I've tested only on the machine that I build,
>>>>and it has one Gigabyte of RAM. I'll try again on a Windows 98 machine,
>>>>and an XP box.
>>>>
>>>>-Mike
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>>From: Alex Burger <[EMAIL PROTECTED]>
>>>>>Date: 2004/09/22 Wed PM 09:32:58 EDT
>>>>>To: [EMAIL PROTECTED]
>>>>>CC: [EMAIL PROTECTED]
>>>>>Subject: Re: patches are applied
>>>>>
>>>>>
>>>>>I'm using 6.0 for the building, like usual. Nothing else has really
>>>>>changed on my system except for some MS updates. Did you try my binary
>>>>>or are you only testing your compiled code?
>>>>>
>>>>>Alex
>>>>>
>>>>>[EMAIL PROTECTED] wrote:
>>>>>
>>>>>
>>>>>
>>>>>>HI, Alex. Are you building with MSVC++ 6.0 or 7.0 ?
>>>>>>I use 6.0 with out problem...
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>From: Alex Burger <[EMAIL PROTECTED]>
>>>>>>>Date: 2004/09/22 Wed PM 06:00:53 EDT
>>>>>>>To: [EMAIL PROTECTED], [EMAIL PROTECTED]
>>>>>>>Subject: Re: patches are applied
>>>>>>>
>>>>>>>Hi!
>>>>>>>
>>>>>>>As soon as the public CVS is up to date with the regular one, I will
>>>>>>>build a new binary. It probably will be in a few hours.
>>>>>>>
>>>>>>>Did either of you try the last binary I made? When using that binary, I
>>>>>>>can reproduce the snmpwalk problem on three machines. The only thing in
>>>>>>>common between the machines I can think of is that they are all Win2000
>>>>>>>(one server, two prof) with SP4.
>>>>>>>
>>>>>>>Try doing a walk of .1 and see if it does. I've been testing with a
>>>>>>>snmp v1 walk.
>>>>>>>
>>>>>>>Alex
>>>>>>>
>>>>>
>>
>>
>>
>>-------------------------------------------------------
>>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
>>Project Admins to receive an Apple iPod Mini FREE for your judgement on
>>who ports your project to Linux PPC the best. Sponsored by IBM.
>>Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
>>_______________________________________________
>>Net-snmp-users mailing list
>>[EMAIL PROTECTED]
>>Please see the following page to unsubscribe or change other options:
>>https://lists.sourceforge.net/lists/listinfo/net-snmp-users
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
> Project Admins to receive an Apple iPod Mini FREE for your judgement on
> who ports your project to Linux PPC the best. Sponsored by IBM.
> Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
> _______________________________________________
> Net-snmp-users mailing list
> [EMAIL PROTECTED]
> Please see the following page to unsubscribe or change other options:
> https://lists.sourceforge.net/lists/listinfo/net-snmp-usersN¬HS^µéšŠX¬²š'[EMAIL
> PROTECTED]
> P1„ê'z‡õïCë¢7œ´fŠ{-¢·œz+Þjp)¦W¢>‡LŠx…DA ¢¼¨º¸îv¦z{hŸh¦Ší³*.®šè?ç-¶‚âžìO<+ay·¬µ*hžÊ+yÖò
>
> y§eŠw’z›vàj!z·¡¶ÚþÇçzßé¥Ç(ž×¬¶˜i5ëlžjn±ê왨¥Šx%ŠËMzÛ'š›¬z»%ŠËl²‹«qç讧zÓåy« ±ç…çè–Z0Šx)j¢éì¹» ®&Þ¢·!jx ¢Ø^®ŠmŠ‰ì†Ûi³ÿåŠËl²‹«qç讧zßåŠËlþX¬¶)ߣùÞ¶Éæ¦ë rs==
