Hi all,
I beleive that SNMP v2c has quite robust way of reporting errors. It
has 13 error codes. But does net-snmp has support for handling these error
codes. Whenever an "snmpget" command is issued, the command returns "0" or
"1" depending upon success or failure of the command. But how do
An unsigned variable should remain unsigned regardless of the bit length :)
-david
Scott ~ # diff -ruN /usr/include/net-snmp/types.h~
/usr/include/net-snmp/types.h
--- /usr/include/net-snmp/types.h~ 2007-01-23 16:07:03.0 -0500
+++ /usr/include/net-snmp/types.h 2007-01-23 16:07
> "c" == cnelson <[EMAIL PROTECTED]> writes:
c> Is there a function or a #define or something that will let a subagent
c> determine what configuration file the agent is using? I found
c> get_configuration_directory() but it returns a path of several
c> directories. What I want is something
> "DS" == Dave Shield <[EMAIL PROTECTED]> writes:
>> snmpget -v 3 -l noAuthNoPriv localhost sysUpTime.0
DS> Unfortunately, the FAQ is wrong :-(
DS> [You just can't get the staff these days]
DS> You can have unauthenticated SNMPv3 requests, but you
DS> do still need to provide a (valid) u
Is there a function or a #define or something that will let a subagent
determine what configuration file the agent is using? I found
get_configuration_directory() but it returns a path of several
directories. What I want is something that returns, for example,
/usr/share/snmp/snmpd.conf.
---
Dave,
> SNMP is basically a point-to-point protocol.
At least in IPv4 SNMP broadcasts were a quite common way to discover devices
in the network and then manage them using point-to-point SNMP GETs and SETs.
Even snmpd implements this functionality for IPv4 (listening on udp:161).
So can I assu
On 23/01/07, Fabricio Tofoli <[EMAIL PROTECTED]> wrote:
> The problem is that snmpd is not responding. Is snmpd
> listening to multicasts to ff02::1?
I doubt it.
SNMP is basically a point-to-point protocol.
It's not really designed for multicast operation.
Unless you've explicitly hacked the agent
Hello all,
I'm trying to use snmpd to test my SNMP discovery code.
Basically my module sends an multicast SNMPv1 packet to the ff02::1 address.
The problem is that snmpd is not responding. Is snmpd listening to multicasts
to ff02::1?
Thanks,
Fabricio
OK, so I misread the patch details...
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
Michael Kirkham
Sent: 23 January 2007 11:45
To: net-snmp-users@lists.sourceforge.net
Subject: Re: Encrypting trapsess entries in snmpd.conf
I'd say bug, because it al
I think that there's still a potential bug here though.
create_user_from_session sets the storage type to read-only, but chances
are that the user was previously not read-only if it was read from
persistent config.
I don't know if this would then prevent the usm details being changed
for the user v
Question about Trap generator, 1) applied by different process in appli in
parrallel (multithreads), 2) following send message (effect of ACK or Timeout)
1 - If a manager application has different processus requiring to send eaxh
trap, does the application need to enque the call to the trap genr
Having been digging for bugs in the target/notification/filter table
mechanisms, I can confirm this.
When it's time to send notifications to targets registered via the MIB
tables, the agent calls send_notifications (snmpNotifyTagble.c).
send_notifications calls get_target_sessions (target.c) to
On Tue, 23 Jan 2007, Michael Kirkham wrote:
> Date: Tue, 23 Jan 2007 03:44:56 -0800 (PST)
> From: Michael Kirkham <[EMAIL PROTECTED]>
> To: net-snmp-users@lists.sourceforge.net
> Subject: Re: Encrypting trapsess entries in snmpd.conf
>
>
> I'd say bug, because it also causes users created via SN
I'd say bug, because it also causes users created via SNMP to receive
SNMPv3 informs via the target/notification MIBs to be deleted on reboot,
and readOnly (according to the StorageType TC) should be saved to
persistent storage, unlike nonVolatile. I submitted a patch (1642071)
about this yes
One other possibility: split targetAddrTable_struct's session pointer into
one for traps and one for informs, configure the trap one with the agent's
engineID, and add whichever one (or both) to the result that's
appropriate. Needs to still not create multiple sessions for the same
notificati
Hi,
Could you help me in combining trap subagent and normal parameter subagent
into one subagent binary. As we cant use threads in snmp applications.
Thanks and regards,
Subrato
-
Looking for earth-friendly autos?
Browse Top Cars by "Green Rating" at
On 23/01/07, Subrato Namata <[EMAIL PROTECTED]> wrote:
> Could you help me in combining trap subagent and normal parameter
> subagent into one subagent binary.
What do you mean by that?
If you have a single AgentX subagent application,
it can register MIB handlers to deal with SNMP requests,
and
17 matches
Mail list logo
