Loading a particular MIB causes init_snmp to crash
Hi, I am using net-snmp v5.5. On loading a particular MIB (attached with the mail), my program crashes on init_snmp API. On removing this MIB from the MIB directory, everything works fine. Please share your views on what may be the issue in this case. Thanks and Regards Gaurav Chaturvedi CLARIION-MIB Description: Binary data -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
SNMP demo app
Hi! In your demo application (http://www.net-snmp.org/tutorial/tutorial-5/toolkit/demoapp/snmpdemoapp.c) you fill the field session.peername using strdup function. init_snmp(snmpdemoapp); /* * Initialize a session that defines who we're going to talk to */ snmp_sess_init( session ); /* set up defaults */ session.peername = strdup(test.net-snmp.org); /* set up the authentication parameters for talking to the server */ Where do you release memory which were allocated by this function? Regards, Sergei -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Loading a particular MIB causes init_snmp to crash
On 10 August 2010 08:14, Gaurav Chaturvedi gauravchaturve...@gmail.com wrote:
I am using net-snmp v5.5. On loading a particular MIB (attached with the
mail), my program crashes on init_snmp API.
On removing this MIB from the MIB directory, everything works fine.
Please share your views on what may be the issue in this case.
Well the obvious error in this MIB is in the TRAP-TYPE definition:
EventMonitorTrap TRAP-TYPE
ENTERPRISE EMC CLARiiON Advanced Storage Solutions
VARIABLES { .}
DESCRIPTION
::= 2
The ENTERPRISES clause is meant to contain the object identifier (name
or numeric OID) of the defining enterprise. It's not an arbitrary string.
See RFC1215, section 2.2.1 for a typical example.
That's almost certainly the cause of the crash.
There are some other errors as well
(OBJECT-TYPE should probably be imported from RFC1212 rather than 1155,
TRAP-TYPE is not imported at all, and the EventMonitorTrap name is
invalid - it needs to start with a lower-case letter)
but I doubt those are likely to cause any real problems.
It'll be the bogus enterprise value that's hitting you - try replacing
it with 'clariion'
and see if that works any better.
Dave
--
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev
___
Net-snmp-users mailing list
Net-snmp-users@lists.sourceforge.net
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Loading a particular MIB causes init_snmp to crash
Hi,
Thanks for the response. I know the error in MIB file (same as specified by
Dave), but should an erroneous value in MIB crash the application?
I want to know some way that can report error on MIB and exit/write a log.
Thanks and Regards
Gaurav Chaturvedi
(+91) 957-944-3976
On Tue, Aug 10, 2010 at 1:29 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote:
On 10 August 2010 08:14, Gaurav Chaturvedi gauravchaturve...@gmail.com
wrote:
I am using net-snmp v5.5. On loading a particular MIB (attached with the
mail), my program crashes on init_snmp API.
On removing this MIB from the MIB directory, everything works fine.
Please share your views on what may be the issue in this case.
Well the obvious error in this MIB is in the TRAP-TYPE definition:
EventMonitorTrap TRAP-TYPE
ENTERPRISE EMC CLARiiON Advanced Storage Solutions
VARIABLES { .}
DESCRIPTION
::= 2
The ENTERPRISES clause is meant to contain the object identifier (name
or numeric OID) of the defining enterprise. It's not an arbitrary string.
See RFC1215, section 2.2.1 for a typical example.
That's almost certainly the cause of the crash.
There are some other errors as well
(OBJECT-TYPE should probably be imported from RFC1212 rather than 1155,
TRAP-TYPE is not imported at all, and the EventMonitorTrap name is
invalid - it needs to start with a lower-case letter)
but I doubt those are likely to cause any real problems.
It'll be the bogus enterprise value that's hitting you - try replacing
it with 'clariion'
and see if that works any better.
Dave
--
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev ___
Net-snmp-users mailing list
Net-snmp-users@lists.sourceforge.net
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Loading a particular MIB causes init_snmp to crash
On 10 August 2010 09:25, Gaurav Chaturvedi gauravchaturve...@gmail.com wrote: Thanks for the response. I know the error in MIB file (same as specified by Dave), but should an erroneous value in MIB crash the application? No - of course it shouldn't. The MIB parser wasn't written in the expectation of such a broken MIB, and really needs to be tightened up. I want to know some way that can report error on MIB and exit/write a log. The first step is to investigate what exactly is causing the parser to crash, and fix that. If you can provide a suitable patch, then we would be happy to include it in the code for future releases. Dave -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Providing an ifTable through a subagent
Hi all. I am taking my very first steps in subagent development. I have developed a subagent which provides the ifTable to a master agent that already provides its own ifTable. So, I edited /etc/sysconfig/snmpd.options and added these: -Dmib_init -I -ifTable I can now see that snmpd does not provide its own ifTable any longer but when I run my subagent, despite successfully connecting to the master, I can not read the ifTable provided by the subagent. My question: Does the -I -ifTable option prevents access to the ifTable even if provided by a subagent? I have followed the mfd tutorial steps, as described here: http://www.net-snmp.org/tutorial/tutorial-5/toolkit/mfd/if-mib/ifTable/index.html Thank you very much for your help! My apologies if I was not clear enough. Just tell me to try rephrasing it again... Regards, Stavros -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Loading a particular MIB causes init_snmp to crash
Hi Dave, Sure, I am trying to fix this. Will provide the same soon. In the meantime, if anybody is able to provide a fix/work-around for this, it will be highly appreciated. Thanks and Regards Gaurav Chaturvedi On Tue, Aug 10, 2010 at 2:21 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 10 August 2010 09:25, Gaurav Chaturvedi gauravchaturve...@gmail.com wrote: Thanks for the response. I know the error in MIB file (same as specified by Dave), but should an erroneous value in MIB crash the application? No - of course it shouldn't. The MIB parser wasn't written in the expectation of such a broken MIB, and really needs to be tightened up. I want to know some way that can report error on MIB and exit/write a log. The first step is to investigate what exactly is causing the parser to crash, and fix that. If you can provide a suitable patch, then we would be happy to include it in the code for future releases. Dave -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Providing an ifTable through a subagent
On 10 August 2010 10:07, Tsolakos Stavros stsola...@gmail.com wrote: Does the -I -ifTable option prevents access to the ifTable even if provided by a subagent? No. The option '-I-ifTable' when given to the main agent simply means don't invoke the routine 'init_ifTable' The effect of this is that the main agent won't try to register the ifTable, leaving the way clear for your subagent to provide this information. The problem will lie elsewhere - either in how your subagent registers with the master agent, or in how it responds to queries for this table. I would suggest that the ifTable is a fairly complicated table to start with. It would probably be a good idea to try with a simpler demo table first, and make sure you can register and query that successfully, before moving on to more compilcated tables. Dave -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Wrong Type (should be INTEGER): NULL
HI,
I have implemented an AgentX subagent.
The mib is standard Q-BRIDGE_MIB, node is
dot1qVlanVersionNumber OBJECT-TYPE
SYNTAX INTEGER {
version1(1)
}
MAX-ACCESS read-only
STATUS current
DESCRIPTION
The version number of IEEE 802.1Q that this device
supports.
REFERENCE
IEEE 802.1Q/D11 Section 12.10.1.1
::= { dot1qBase 1 }
I have generated the code using mib2c dot1qVlanVersionNumber
1.net-snmp style
2. Both ways scalar and tied integer.
code generated and modified from mib2c.scalar.conf is
long dot1qVlanVersionNumber = 1; /* XXX: set default value */
/*
* Our initialization routine, called automatically by the agent
* (Note that the function name must match init_FILENAME())
*/
void
init_dot1qVlanVersionNumber(void)
{
static oid dot1qVlanVersionNumber_oid[] = { 1,3,6,1,2,1,17,7,1,1,1 };
DEBUGMSGTL((dot1qVlanVersionNumber, Initializing\n));
netsnmp_register_scalar(
netsnmp_create_handler_registration(dot1qVlanVersionNumber,
handle_dot1qVlanVersionNumber,
dot1qVlanVersionNumber_oid, OID_LENGTH(dot1qVlanVersionNumber_oid),
HANDLER_CAN_RONLY
));
}
int handle_dot1qVlanVersionNumber(netsnmp_mib_handler *handler,
netsnmp_handler_registration *reginfo,
netsnmp_agent_request_info *reqinfo,
netsnmp_request_info *requests)
{
switch(reqinfo-mode) {
case MODE_GET:
case MODE_GETNEXT:
snmp_set_var_typed_value(requests-requestvb, ASN_INTEGER,
(u_char *) dot1qVlanVersionNumber/* XXX: a pointer to the scalar's data */,
sizeof(long)/* XXX: the length of the data in bytes */);
break;
default:
/* we should never get here, so this is a really bad error */
snmp_log(LOG_ERR, unknown mode (%d) in handle_dot1qVlanVersionNumber\n,
reqinfo-mode );
return SNMP_ERR_GENERR;
}
return SNMP_ERR_NOERROR;
}
When I do snmpwalk I get
snmpwalk -v3 -u xyz -A 12345678 -X 12345678 -a MD5 -l authNoPriv localhost
Q-BRIDGE-MIB::dot1qVlanVersionNumber
Q-BRIDGE-MIB::dot1qVlanVersionNumber.0 = INTEGER: version1(1)
Q-BRIDGE-MIB::dot1qVlanVersionNumber.0 = Wrong Type (should be INTEGER):
NULL
Error: OID not increasing: Q-BRIDGE-MIB::dot1qVlanVersionNumber.0
= Q-BRIDGE-MIB::dot1qVlanVersionNumber.0
snmpget is working fine
snmpget -v3 -u xyz -A 12345678 -X 12345678 -a MD5 -l authNoPriv localhost
Q-BRIDGE-MIB::dot1qVlanVersionNumber.0
Q-BRIDGE-MIB::dot1qVlanVersionNumber.0 = INTEGER: version1(1)
This behaviour is same when code is generated using tied integer.
Why I am getting Q-BRIDGE-MIB::dot1qVlanVersionNumber.0 = Wrong Type (should
be INTEGER): NULL
Error: OID not increasing: Q-BRIDGE-MIB::dot1qVlanVersionNumber.0
= Q-BRIDGE-MIB::dot1qVlanVersionNumber.0
in snmpwalk output?
I searched for this in archive and found that this is probably due to my
agent is broken.
Can you please help me out in figuring Where is agent broken and how to fix
it?
Thanks in advance!
--
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev ___
Net-snmp-users mailing list
Net-snmp-users@lists.sourceforge.net
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Re: net-snmp init_XXX() get data periodically( or cyclic ) problem
Hi - For the -1, I was referring to the value in the .h file. For example, #define LLDPLOCMANADDRTABLE_TIMEOUT 60 Change the 60 to -1. The mib2c with the cache option generates a default value (in this case, 60) and you can adjust it. I haven't tried the on the fly option yet. There is another mib2c for scalars - mib2c.scalar.conf. Kathy McLeod Alexander King new...@foxmail.c omTo Kathy McLeod/Rochester/i...@ibmus, 08/06/2010 04:52 net-snmp-users AMnet-snmp-us...@lists.sourceforge.n et cc Subject Re: Re: net-snmp init_XXX() get data periodically( or cyclic ) problem Thank you all kind man! I will do my best get a prototype the way Dave teaching me ASAP. Also, I thought -1 was the value for load every time. I saw the letter later Dave's wrote back,is it the meaning use mib2c -S cache=-1 -c mib2c.iterate.conf uit can get the really data on the fly???and how about the saclars data? I got a little inspiration,I think. -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: OCTET STRING Size -SNMP-V3
The maximum length of an OCTET STRING is 65535. However, it's unlikely you want to send that much data over UDP, however. -- Wes Hardaker Cobham Analytic Solutions -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: implementing traps...
On Thu, 5 Aug 2010 14:01:21 +0530 (IST), ganeshyell...@bel.co.in said: g after this what to edit in coldStart.c? You'll need to call the created function from within your code where you want the trap to be sent. The coldStart.c file is likely mostly complete unless you want to add additional variables, like the comments talk about. -- Wes Hardaker Cobham Analytic Solutions -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
SNMP Make error
Hi, While running: make test, I get the following error. Thanks in advance for any help you provide. make test making all in /usr/local/src/net-snmp/net-snmp-5.5/snmplib making all in /usr/local/src/net-snmp/net-snmp-5.5/agent /bin/bash ../libtool --mode=compile gcc -I../include -I. -I../agent -I../agent/mibgroup -I../snmplib -I/usr/local/ssl/include -g -O2 -Usolaris2 -Dsolaris2=solaris2 -c -o snmp_vars.lo snmp_vars.c libtool: compile: gcc -I../include -I. -I../agent -I../agent/mibgroup -I../snmplib -I/usr/local/ssl/include -g -O2 -Usolaris2 -Dsolaris2=solaris2 -c snmp_vars.c -fPIC -DPIC -o .libs/snmp_vars.o In file included from /usr/include/kvm.h:14, from kernel.h:11, from snmp_vars.c:143: /usr/include/sys/proc.h:202: error: syntax error before p_t1_lgrpid /usr/include/sys/proc.h:203: error: syntax error before p_tr_lgrpid /usr/include/sys/proc.h:337: error: syntax error before '}' token In file included from kernel.h:11, from snmp_vars.c:143: /usr/include/kvm.h:44: error: syntax error before '*' token /usr/include/kvm.h:45: error: syntax error before '*' token /usr/include/kvm.h:48: error: syntax error before proc_t *** Error code 1 make: Fatal error: Command failed for target `snmp_vars.lo' Current working directory /usr/local/src/net-snmp/net-snmp-5.5/agent *** Error code 1 The following command caused the error: if test snmplib agent apps man local mibs != ; then \ it=snmplib agent apps man local mibs ; \ for i in $it ; do \ echo making all in `pwd`/$i; \ ( cd $i ; make ) ; \ if test $? != 0 ; then \ exit 1 ; \ fi \ done \ fi make: Fatal error: Command failed for target `subdirs' Regards, Ernie -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
RE: SNMP Make error
From: Ernie Lacson [mailto:elac...@successfactors.com] Sent: Tuesday, August 10, 2010 2:12 PM [SNIP] In file included from /usr/include/kvm.h:14, from kernel.h:11, from snmp_vars.c:143: /usr/include/sys/proc.h:202: error: syntax error before p_t1_lgrpid /usr/include/sys/proc.h:203: error: syntax error before p_tr_lgrpid /usr/include/sys/proc.h:337: error: syntax error before '}' token [/SNIP] Looks like the problem is in your include headers. Try compiling a minimal app that includes kvm.h - that should fail as well. HTH, Mike -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
What is this OID?a very strange,long OID problem.
Hi, Like previous mails,I got the generated code use mib2c -S cache=-1 -c mib2c.iterate.conf ,but I got NetsnmpCachefree type comflict problem,in my uit.c file, I deleted the int return type keyword before the XXXcachefree function, then the compiling wrong cleared, it make me very happly. and then I run make install command got snmpd snmptrapd executable program and run it in my device,every time when I use MG-soft walk the whole MIB-tree,the snmpd corrupted when the browser run the OID( something like the attachment,the red line in the pic). I cannot find where my code is wrong,and what the OID's value is. Help me ! alexattachment: Snap1.jpg -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
