_synch
_response__
raise GetError, errmsg
[0]=yapsnmp.GetError
[1]=couldn't obtain response
Any ideas? Could this be achieved more elegantly without going out to an
external script?
Thanks
dave
Send instant messages to your online friends http://uk.messenger.yahoo.com
and it could farm out pass requests itself.
The approach I've taken to solve my original problem is a different one, but it
would be handy, I think, to have a re-entrant snmp agent...
Comments?
dave
Send instant messages to your online friends http://uk.messenger.yahoo.com
ing the (five-years-old) wheel.
If there genuinely isn't an implementation I'd be happy to submit my script.
cheers
dave
Send instant messages to your online friends http://uk.messenger.yahoo.com
---
SF email is sponsored by - T
--- Robert Story <[EMAIL PROTECTED]> wrote:
> On Tue, 1 Mar 2005 18:10:33 +0000 (GMT) Dave wrote:
> D> So my python script handles the get() and getnext() by talking back to the
> D> agent and filtering out unwanted OIDs, then returning the result back to
> the
> D> c
the program
will be accepted and made available to net-snmp users before they give me the
go-ahead, so any feedback would be very welcome.
Best Regards
dave
Send instant messages to your online friends http://uk.messenger.yahoo.com
---
SF e
ill up the log when I specify -DALL.
I've been hitting my head against this for two days now. I hope it's
painfully obvious.
Thanks for a great package and thanks for your help!
Dave
CLIENT COMMAND:
./snmpwalk -M /root/mibs -c rocommunity -v 2c -r 0 -d localhost -DALL
Note:
n hitting my head against this for two days now. I hope it's
> painfully obvious.
>
> Thanks for a great package and thanks for your help!
>
> Dave
>
>
> CLIENT COMMAND:
>
> ./snmpwalk -M /root/mibs -c rocommunity -v 2c -r 0 -d localhost -DALL
>
> Note: I hav
"sess_process_package" reqid seems right, the message id is always 0.
>> But
>> that may just be because the server will assign it. what look like
>> loopback tests fill up the log when I specify -DALL.
>>
>> I've been hitting my head against this fo
t -v 2c -c rocommunity 172.17.4.3 sysUpTime.0
>
> Have you configured the agent to recognise/accept the community
> name "rocommunity" ?
>You'd need a line such as
>
>rocommunity rocommunity
>
> in the snmpd.conf file.
>
> Dave
>
bmoninsp_trap");
> SOCK_CLEANUP;
>
> That way you'd only have one session open all the time,
> and wouldn't need a separate socket for each call.
>
> Dave
>
> --
> Better
Hi,
I am trying to set up disman restarts on a process. I am pretty sure I
don't have
acm set up properly, and as a result of this I see a memory leak every
time I attempt,
apparently, to run the daemon with the line that is commented out below
enabled.
When I run start the daemon with a -d op
Hi,
Thanks for your response. I've answered your questions as well as I could
and included a "watch" showing the memory leak.
> Which version of the agent are you using?
>
I am compiling from the package 5.6.2 pre 1
> Is this an agent that you compiled yourself, or a vendor-supplied version?
>
ne (causing the problem) is missing the "NAME"
> field.
> So it's probably taking "prCount.3" as the name of the entry, and "< 1" as
> the expression to monitor.
>
> I'm a little surpr
s just a cover for an access violation,
i.e., but that doesn't really explain why the type is wrong on the disman
read.
Thanks,
Dave
--
Live Security Virtual Conference
Exclusive live event will cover all the
The command still isn't run automatically.
I worked it through with snmpset, and it seems to work, once, then
the snmpset commands timeout.
Then this happens:
[dave@localhost net-snmp-5.6.2.pre1]$ apps/snmpset -c dismanrw -M mibs -m
ALL -v 2c 172.17.4.3:1161 prErrFix.2 = 0
UCD-SNM
Which is the best (v5) model for this kind of dual implementation in a
subagent?
What is the (social) procedure to follow if I want to add new DataTypes,
i.e., (extend the Entity Sensor MIB)?
Thanks,
Dave
--
Keep
tion
rule (RFC1902 sec. 10 or thereabouts), it seems I can "append" but not
"reorder", but must there then be community wide follow through?
I'm presuming and ready to argue that it would be a MISTAKE to put out a
MIB with Henrys(13) in it, but is there a better way to do it t
6.2pre1.
Any ideas? Please keep in mind that the version with three arguments at
the top of this posting ran just fine for a long time, but adding that
triplet of arguments to the end OR using an index >= 3 breaks it. We have
adjusted the MIB and even restarted the Nag
ith three arguments at
> the top of this posting ran just fine for a long time, but adding that
> triplet of arguments to the end OR using an index >= 3 breaks it. We have
> adjusted the MIB and even restarted the Nagios box.
>
> Thank you very much!
>
> Dave
>
>
>
> > It looks suspiciously as if it's trying to run as an
> > AgentX subagent - is that expected?
>
> Yes.
OK - Are you running an AgentX master agent?
If so - where is it listening for connections?
Dave
PS:
s a subagent.
Now you say that you are running the Net-SNMP as a *master* agent,
and the subagent is embedded in your instrumentation.
So I will ask again.
Which application is giving this error - the main snmpd program
or your instrumentation?
What settings have you got in the 'snmpd.conf&#
a source distribution, or as part of the
documentation, etc. accompanying a binary distribution.
This holds whether or not you have made any changes yourself.
Dave
---
This SF.Net email is sponsored by: Oracle 10g
Get certified on the
:
- create a group containing the user 'wes'
- create a suitable view
(if one doesn't already exist)
- allow access for that group to that view
The simplest approach would probably be to add the user 'wes'
to
> What should I do to see all the MIB-2 tree? I did not modify snmpd.conf.
Modify snmpd.conf:-)
See the FAQ entry
I can see the system group, but nothing else. Why?
Dave
---
This SF.Net email is sponsored by: Oracle 10g
ess to my snmp tree.
What do you mean by "everybody"?
More importantly, are you talking about SNMPv1/2c or SNMPv3 ?
> Could you send a sample basic working snmpd.conf?
See the FAQ entry
How do I configure access control?
That's about as simple as it can get.
Dave
ilar) command line.
Anything related to the operation of the agent (like the directives
listed in 'snmpd -H') should indeed go in 'snmpd.conf'
Dave
---
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hott
without the '.c' suffix)
Dave
---
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g.
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads
> I'm using 5.1.1 but haven't gotten any patches.
> Is this something in a recent change to configure?
No - it's been like this since 10th March 1997 (UCD v3.2)
Dave
---
This SF.Net email is sponsored by: Oracle 10g
nmp_watcher_info *winfo =
netsnmp_create_watcher_info(
myStringObject, strlen(myStringObject),
ASN_OCTET_STR, WATCHER_FIXED_SIZE);
netsnmp_register_watched_scalar( reg, winfo );
Dave
---
This SF.Net
_registration() call?
Remember that this is registering a scalar object,
so you'll need to add the instance subidentifier.
Try
snmpget -v2c -c public localhost 1.3.6.1.4.1.10.1.2.0
Dave
---
This SF.Net email is sponsored by: Oracle
Dave> Have a look at "netsnmp_register_watched_scalar"
Dave> Not only does this cope with string values as well as integer ones,
Dave> it also handles scalar objects correctly (which netsnmp_register_instance
Dave> doesn't - in particular requests for an invalid ins
the source address filtering.
But other than that, it shouldn't be too difficult to implement.
Dave
---
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g.
Take an Oracl
> better (but not so elegant) to use the INTEGER data type instead.
Why not have a look at the (SMIv2) BITS type.
This *is* valid in SNMP, and should be supported by other toolkits
as well.
Dave
PS: Please - no HTML mail. Thanks
---
This
ade by the programmer.
The agent doesn't know (or care) that these tables are connected.
Dave
---
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g.
Take an Oracle 10g cla
et_oid, target_oid_len,
upperB_oid, upperB_oid_len) < 0 )) {
printf("target lies strictly within the desired range\n");
}
Replace ">" and "<" with ">=" and "<=" if you want to include the
boundary OIDs as well.
Dave
--
snmpset . UCD-SNMP-MIB::extErrFix.{N} i 1
(where {N} is the appropriate index) should reboot the system.
But be *VERY* careful about your access control settings.
This *NOT* something you want to allow general access to with
the "public" or "private" communities. (Or p
ned, the code is basically:
printf( "IP Address is %d.%d.%d.%d\n",
vb->val.string[0],
vb->val.string[1],
vb->val.string[2],
vb->val.string[3] );
Dave
---
This SF.N
result
in the same answer.
> They know that Bill and William are the same person but
> I don't have to know that to find you.
Correct.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux,
> Let's try a slightly more confusing analogy:
> Bill has cable, Dave has satellite, and Bob has an antenna.
>
> Each tunes-in to Are You Being Served? But:
> Bill's on channel 4, Dave's on 57, and Bob's on 17.
That's a somewhat misleading example.
more links, all referring to a
single underlying file. There will be two or more valid
names for this file, but the contents are the same for all
of them.
Dave
This message was sent usin
uot;Makefiles.top" thusly:
< INSTALL = $(LIBTOOL) --mode=install @INSTALL@
--
> INSTALL = $(LIBTOOL) -s --mode=install @INSTALL@
and re-running configure.
[Not actually tested, I hasten to add]
Dave
---
This SF.
.g. via the "sysname"
directive, or SET requests).
It certainly doesn't look as if it tracks changing values of the
underlying hostname.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to
ything interesting happens.
If you don't have the documentation installed on your system, it might
be worth doing so. Calls such as "select" are part of the underlying
operating system. They're not specific to the Net-SNMP suite.
Dave
--
ing 'strlen' when creating the 'winfo' structure,
then you must use 'sizeof' here. Or vice versa.
You need to specify both sizes - it doesn't matter which order.
But you *must* declare this as a MAX size, not a FIXED size watcher.
Dave
-
is configured.
(At a guess, it's probably '1', but it might not be).
Try
snmpwalk UCD-SNMP-MIB::extTable
and this should show what value to use for N.
> And as soon as we do the set will it restart th
d_process to figure this out and I
> shouldn't fix what's not broke.
I'd say it's not really broken, and isn't worth fixing.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Li
for ( request = requests; request; request=request->next_variable)
loop (and with with 'request' not 'requests' internally).
That's the most likely cause that springs to mind immediately.
Dave
---
and.
It's also possible to configure an initial value
using the "sysname" directive in snmpd.conf.
If either of these have happened, then it probably
wouldn't be appropriate to track changes to the
underlying hostname.
Dave
ropriate step for RedHat to
have taken, and we will not contemplate including this patch within
the Net-SNMP distribution.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield
> Dose the net-snmp and ucd-snmp have source code that I can
> development them on the windows OS? Thanks a lot!!
Yes.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers
er starting xinetd :(
Please note that 'snmpd' is run as a standalone daemon, and is *not*
typically under the control of xinetd.
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, s
ng on the exact order in which the varbinds are processed),
but this list is the safest and most reliable.
> And module is not loaded. Why?
The module is *only* loaded by step 'd)', when you explicitly
ask for this to be done.
Dave
-
whether this is an AgentX
subagent (and hence pass the trap on to the master agent), or a
master agent (and hence send the trap directly to the registered
managers).
The MIB module code does not need to worry about this.
Dave
---
This SF.Net em
failure.
That was one of the differences introduced between SNMPv1 and SNMPv2.
Missing values are not longer regarded as failures - just missing values.
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The wo
cter values are.
I don't understand where the problem lies?
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #1 Open Source Desktop Event.
GNOME Users and Developers European Confer
er*
rather than a full OID.
Try running configure using:
--with-enterprise-oid=18293
You should find that things compile OK.
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #1 Open Source De
ctet string that is sent in the second request?
(Run the command with '-d' to dump the outgoing packet,
or '-Ddump' to display a formatted version).
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Uni
o far. It would be very useful to hear about whether
this works for you - particularly on any {Free,Net,Open}BSD systems
you might have.
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #
d identify where the source of the problem lies.
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #1 Open Source Desktop Event.
GNOME Users and Developers European Conference, 28-30t
5.1.2 code as well,
or would this count as a "new feature" ?
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #1 Open Source Desktop Event.
GNOME Users and Developers European Con
's the name of the code file?
Do they match?
Have a look at the file 'agent/mibgroup/mib_module_includes.h'
Does this contain the name of your module header file?
Have a look at the file 'agent/mibgroup/mib_module_inits.h'
Does this contain the name of your initial
nding on how the agent was compiled)
Dave
---
This SF.Net email is sponsored by: GNOME Foundation
Hackers Unite! GUADEC: The world's #1 Open Source Desktop Event.
GNOME Users and Developers European Conference, 28-30th June i
hich seemed to be what Mike was assuming. (Unless I misunderstood
the message).
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-aut
AX SnmpAdminString
(or similar) - so that it would use the DISPLAY-HINT "255a"
Dave
---
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-a
.1.1.1), while, I should have got the value for oid x.1.2.1.1
Correct - assuming B is accessible.
(Otherwise you should get x.1.3.1.1 again)
You should certainly *not* get x.1.1.1.1
Dave
---
This SF.Net email is sponsored by the new Insta
#x27;s only in the main CVS tree, so won't be included in the
upcoming 5.1.2 release (unless there's a general consensus that it should)
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the expe
eing defined directly under the
'netSnmpExamples' subtree, rather than using the relevant
portions of the wider framework.
That works equally as well - it's just a slightly different
approach.
There's no One Right Way to do this.
Choose something that makes sense to you, an
.sh -g .1.3.6.1.4.1.96.255.2.0 ; echo $?
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center i
Dave> It looks as if the main CVS branch has rather more robust argument
Dave> parsing code, which *does* report on unrecognised options.
Dave> Is it worth applying this approach to the 5.1.2 code as well,
Dave> or would this count as a "new feature" ?
Wes> Though I agr
*iinfo =
SNMP_MALLOC_TYPEDEF(netsnmp_iterator_info);
+ iinfo->flags |= NETSNMP_ITERATOR_FLAG_SORTED;
It's still not the most efficient helper, but it should help.
There's certainly no need to re-implement this feature.
Dave
. I hope to get back to this again
over the next month (and it may have been fixed in the meantime),
but the simplest approach is definitely to use the internal
RowStatus handling, if this is appropriate for your MIB table.
Dave
---
This SF.
dDemoPublic examples/example"
What happens if you try with a bare "configure" - with no extra options?
I know that's not what you need to use, but it would be useful to know
whether you get the same warnings/errors.
Dave
er variable,
and declare them using ASN_COUNTER64
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Ce
27;re linking with the '-lcrypto' library.
Is OpenSSL installed on your system?
How did you run "configure" ?
Dave
PS: Please - no HTML mail. Thanks
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from
first RH9 CD). That includes the missing library.
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone
> Somebody can help me about how to find documentation regarding
> SNMPv3 support by UCD-SNMP agent?
Have you read
README.snmpv3
and the man pages 'snmpd.conf(5)' and 'snmpcmd(1)' ?
Dave
--
Quoting Anuradha Bhakta <[EMAIL PROTECTED]>:
> I would like to populate snmpNotifyTable(RFC 3014) though the
> configuration file(snmd.conf). I did not find any directive that has to
> be added to this snmpd.conf
Try "trapsink&q
.so
As Robert said, what's missing is a link from 'libelf.so' to
'libelf-0.76.so'.
Try:
ln -s libelf-0.76.so n/usr/lib/libelf.so
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn f
're not in a position to switch to the net-snmp v5 releases,
then we'd strongly recommend that you try using 4.2.6.
There's been a significant amount of work done since 4.2.2,
including addressing the POD problem.
Things ought to work better with 4.2.6
Dave
PS: Ple
that you're asking for the wrong OIDs
(as David suggests). Or it may be a problem with
the agent configuration.
Have you read the FAQ?
In particular, see the entry:
I can see the system group, but nothing else. Why?
Dave
---
This SF.N
fer advice. Thanks. ]
> Hi, Dave. Yes, I have read that man page. I have read (right now)
> the README.snmpv3 file too.
>
> An observation: my UCD-SNMP agent was installed by means of a rpm
> file. Can you aswer these two questions?
>
> 1. In the README.s
onf file
that's being read. If you're running a pre-installed version
of the agent, then that will typically be looking in somewhere like
/etc/snmp/snmpd.conf rather than /usr/local/etc/snmp/snmpd.conf
Try deliberately putting an invalid token into the config file
and restarting the agent
> I have done what the FAQ told me which is same as that Robert said. but
> make action doest work yet.
OK - what does
ls -l /usr/lib/libelf*
show?
Dave
---
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
o"
You've got what looks like an object file - "libelf.o"
but that's not the same.
You need a file with the suffix '.so', not '.o'
Dave
---
This SF.Net email sponsored by Black Hat Briefings &am
I have a question on how to run configure: ./configure
> --with-mib-module=host. I was not able to find such util in my
> installation
That's used for compiling from source.
It's part of the source tarball - not part of a pre-compiled distribution.
Dave
---
source) is linking with the RedHat-provided
library (5.0.9).
While this probably ought to work, it is very susceptible to subtle
differences between the two.
It'd be a lot safer to remove the RedHat-provided libraries before
you compile your source tree.
Dave
STARTAGENT and change the option
'-p $SNMP_SNMPD_PID_FILE'
to be
'-P $SNMP_SNMPD_PID_FILE'
I'm not sure why the -p version doesn't work, but that's
the change that seems to have been made for the 5.1.2 pre-release.
Dave
--
> send_easy_trap(SNMP_TRAP_ENTERPRISESPECIFIC, 1);
> but my subagent mean :
> send_trap: Unknown PDU type
Which version of the Net-SNMP suite are you using?
There was a problem with sending traps from subagents,
but the most recent releases (5.0.9, and 5.1.1)
should both work correct
ys about encryption.
If it hasn't found the openSSL libraries, have a look in 'config.log'
which should indicate why not.
Dave
---
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings &a
bug in the AgentX protocol handling.
It should be fixed in 5.1.1.
Please try with that version.
Dave
---
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
d
n cope with watching string buffers as well as integers.
(with a bit of fiddling for writeable variable-size values).
A search on the mailing list archives should give some examples
of how to use this.
Dave
---
This SF.Net email sponsored by Bla
> if net-snmp-5.1.1 has, then how to configure it to work as a SNMP proxy
> agents?
Have you actually tried looking at the documentation?
Like the man page for 'snmpd.conf' ?
There's a whole section there on "proxy".
---
This SF.Net email sp
x27;mib2c' is written in perl, this module is pretty
fundamental.
Try following the instructions in the error message,
and see if that helps.
Dave
PS: Please - no HTML mail. Thanks
---
This SF.Net email sponsored by Black Hat Briefings
whatsoever with what information
the agent will return.
Dave
---
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical exp
d draw pretty graphs of the
traffic over the router, have a look at MRTG or Cricket.
Dave
---
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital sel
tion to fix it.
Although it uses our libraries, it's not really a module
that we know anything about.
Dave
---
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24
> Dave,
>I tried using "netsnmp_register_scalar" for an INTEGER type scalar
> ... and it returned the error No such object instance exists.
>However, if I use netsnmp_register_instance with
> netsnmp_instance_int_handler in netsnmp_create_handler_registrati
s row:
snmpget myhost myColOne.4 myColTwo.4 myColThree.4
Or you can use 'snmpgetnext' to walk through the table a row
at a time:
snmpgetnext myhost myColOne myColTwo myColThree
snmpgetnext myhost myColOne.1 myColTwo.1 myColThree.1
tal checksum for each message.
Both of these stages are one-way - you can't use the signature
of a message to get the localised key, and you can't use the
localised key to get the pass phrase.
Dave
---
This SF.Net email sponso
example2
this/example2_part1.c
that/example2_part2.c
theother/example3_preview.c
-I../../../mibgroup
(all as one line, of course).
Dave
---
This SF.
1 - 100 of 6721 matches
Mail list logo
