Hello KM,
I'm not super tuper expert in net-snmp, but from the trace log I would say:
1) your handler that cares for snmp object is NOT CALLED at all. Instead of
it is
called null handler->>
2005-03-07 15:19:35 handler:calling: calling handler null
2005-03-07 15:19:35 trace: netsnmp
Hello all snmp lovers :))!
I wonder if anybody can explain to me, how handlers are invoked. The only
knowledge I have, I got from debugging net-snmp :(.
Beside that, I've written my own helper handler takes care for returning
values, etc... When handling multiple requests from the same handler
Hello Robert, Wes,... please help!
I wonder how can be some request droped from the registered handler in some
special case, when I don't want to be snmp pdu genereated and returned to
the manager above?
To be a little more specific: I've registered handler 'HandleR' that takes
care for data
Hello,
Talking about multithreads...
I wonder, how can I implement sending traps in agent, triggered from some
other thread that listen independent events?
In other words: I have snmpd thread that reacts on snmpXXX events. Where and
how in the agent I should add code that also sends traps de
Hello,
I read at the:
http://www.net-snmp.org/dev/schedule.html
that new release of net-snmp package will be in July 2005...
I recently developed quite complex snmp system with master and seven
subagents, based on the version 5.2.1.rc2. Time to times I encounter or
better 'invent' error, tha
is planned next version?
Regards,
Igor
> -Original Message-
> From: Grasic Igor [mailto:[EMAIL PROTECTED]
> Sent: 31. avgust 2005 15:28
> To: 'net-snmp-coders@lists.sourceforge.net'
> Subject: When will be next release of net-snmp?
>
>
>
TECTED]
> Sent: 6. september 2005 16:36
> To: net-snmp-coders@lists.sourceforge.net
> Cc: Grasic Igor
> Subject: Re: When will be next release of net-snmp?
>
>
> On Tue, 2005-09-06 at 10:19 -0400, Robert Story wrote:
> > DS> But we seem to be making noises about releasin
Hello,
I know it is almost christmans, everyone probably are way out to holidays...
But. I have problem with dead master agent for a long time from version
5.2.1. My system is configured as follows (embedded linux, kernel
2.4.19-rmk7-pxa1):
- master agent (included modules: agentx/master; exc
Hello, friends!
I do not want to be annoying, but... I need help. Now realy badly. I'm
interested, if anybody discovered something about this issue?
Have a nice day,
Igor Grasic
Igor Grasic, Ms. D.
Development Engineer
Iskra Transmission, d. d., Ljubljana
Stegne 11, SI-1000 Ljub
Hello Dave,
>
> Have you tried 5.3?
>
I have to admit that I did not because it takes me two or three days to
upgrade from one version to another. This 'feature' is present from the
begining of the snmp story for our project; from version 5.1.1 (I wrote
wrong version in first mail). At least
Hello Robert,
I used the pathc you've atached, but... did not help. Master becomes deaf
after a indeterminate of up and down events of the subagent in both
5.2.1.rc2 and 5.2.2 versions. In test configuration I used only one subagent
and of course master. There is also memory leak after shutdow
Hello Thomas!
>Can you please post a gdb backtrace for this crash?
Bellow is bt of crash.
Nice monday,
Igor G.
=
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1085344384 (LWP 3093)]
0x0808371d in _
Hello Dave, Wess!
here is bt of one event of the deaf master agent. It seems to me, that some
agentx session went wrong... and gdb received SIGPIPE. Is that help you any
way of helping me?
Nice day,
Igor Grasic
Program received signal SIGPIPE, Broken pipe.
[Switching to Thread -108450
Hello Robert, Dave, Wes!
Related to my poor deaf or better dead master agent I discovered few things.
When I manage to achive the state of 'darknes' :) at master agent I I
connect remotely to the master with the gdb. There is no other living agentx
client or something else to disturb my test ca
Hello coders!
This mail has already been sent, but in mailing list it is arranged in wrong
order so I'm sending it again.
Nice day,
Igor G.
___
Related to my poor deaf or better dead master agent I discovered
ueue instead of
processed immediatly. Bellow is the listing of from gdb with marked key
points.
So, what is going wrong before that event, that set is processed in wrong
way?
Nice weekend,
Igor G.
> -Original Message-
> From: Robert Story [mailto:[EMAIL PROTECTED]
> Sent: 9. januar
Hello,
I'm newcomer at net-snmp community. In our firm we want to build snmp agent,
responsible for collecting data about status of connected HW modules to the
main platform, running on embedded Linux. Our test system is still running
on RH 9.0. I used mib2c compiler, access_functions template
rking :( in the daemon mode. I also try to comment out
debug sentences in the snmp.conf file, but nothing changed. By the way, at
the begining daemon worked. I personally think, that is somethink wrong in
my code. Does anybody have any idea?
Greetings,
Igor
-----Original Message-
From: Grasic
.conf
(!). Configuration file is not there, snmpd is not complaining neither
work...
Anyway, if anybody want to see *very* simple example of table, do not
hesitate to contact me.
Have a nice day from Slovenia,
Igor
-Original Message-
From: Grasic Igor [mailto:[EMAIL PROTECTED]
Sent: 7.
Hello,
I read I lot of stuff about registering rows in the table by different
subagents, but I did not find any clear implementation, despite that Wes
actualy gave answer to very similar question(bellow).
Anyway, I generated *.c and *.h files with mi2c.iterate_access.conf and
want, that two d
Hello all,
I don't wanna be annoying, but I wrote question on the 8th of July in this
mailing list and nobody answer me. I have very tight scedule I would
appreciate any help, tip that somebody would give me for the problem bellow.
Thanks,
Igor
I read I lot of stuff about registering
Hello everybody,
Design of our system requires one master agent and seven subagents modules
(currently on Linux). I've writen most of the code with the help of
netsnmp_register_instance() to implements particular row from different
subagents and together they represents complete tables of data.
Hello codders, John,
The third way to register some rows in table via AgentX is to use
'netsnmp_register_handler' with set 'range_subid' and 'range_ubound' to the
point in the oid, where is defined column number for the object and with the
oid for the whole (!) row, included with the ifIndex
Hello every body,
I did a lot of playing recently with registering handlers for extended range
of the oids. When I setup up range_subid to the *length* of the registration
oid of the handler (and also setup some upper limit with range _ubound),
appropriate handler is *not called* for range abo
e. Have you tried this?
I think that this wouldn't go, because 'netsnmp_*_table_*' helpers uses
table_helper_handler(), which asume, that you've registered oid of the
*table*.
>Have you seen any other mail or docs about this? Do you have any more
detailed code examples?
table. Me
happy. :)
Prevent instaling headers does not work... :(
Have a nice day,
Igor
P.S.: It's snowing now in Slovenia. :)))
-Original Message-
From: Wes Hardaker [mailto:[EMAIL PROTECTED]
Sent: 12. december 2004 6:25
To: Grasic Igor
Cc: [EMAIL PROTECTED]
Subject: Re: Mem
26 matches
Mail list logo
