Re: MODE_SET_*
On Mon, 2005-10-17 at 18:42 +0200, Roman Medina-Heigl Hernandez wrote: I missed the part that a SET request can set several variables, not only one. Then it is some kind of attempt to reach a pseudo-atomic operation Exactly. (which never gets really true since COMMIT may unexpectedly fail). But there is a responsibility on you (as a MIB developer) to take every care that it does not! If applying an assignment could realistically fail, then you should be doing this during the ACTION pass, and be ready to reverse it in the UNDO pass. If you're simply updating some internal variables - that's probably safe enough to do in the COMMIT pass. And there are some situations where reversing an assignment is likely to be more unreliable than acting on it in the first place. In which case COMMIT is probably the least worst option. But if acting on an assignment can potentially fail, and it's realistic to reverse it, then this should be handled in the ACTION/UNDO passes. Yes, this is more work than COMMIT-based handling, but laziness is no excuse for irresponsibility :-) Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: mib2c
On Mon, 2005-10-17 at 14:43 -0400, Eugen wrote: Thanks for your answer. Actually what I did was the following: - I've downloaded the package net-snmp-5.2.1.2-1.src.rpm from net-snmp web page I though you said you'd installed the RedHat-provided version of Net-SNMP ? - I've generated the binaries with rpm -bb net-snmp.spec The perl modules aren't built by default - you need to explicitly request that they be include (-with perl_modules). See the head of the net-snmp.spec file. So it seems that in 5.2.1.2 the net-snmp-utils is part of the other modules. Not quite - this is a difference between RedHat packaging and our arrangements, not the version of Net-SNMP. The RedHat RPMs distinguish between the setup required for just running the agent (net-snmp), and for using the command-line tools (net-snmp-utils). We don't - our net-snmp RPM covers both. When trying to run mib2c I received that error message than, following the instructions, I've installed the net-snmp-perlmods-5.2.1.2-1 directly from the binary downloaded on the same net-snmp web page (net-snmp-perlmods-5.2.1.2-1.i686.rpm since I didn't found the rpm source of this module). H. The problem with that particular RPM is that it doesn't specify which O/S it was built against. You may have better luck using one of the O/S-specific perlmods RPMs (rh9, fc2, fc4 or el4) Could this be caused by the fact that the binaries generated from the net-snmp source package have all the extension *i386.rpm versus net-snmp-perlmods-5.2.1.2-1.i686.rpm ? Sort of - it's more likely to be the O/S assumptions of that particular RPM don't match your system. That's why the other (i386) RPMs each specify which O/S they were built on. We'll need to ensure that the i686 version requirements are made clearer in the future. And another thing. After the net-snmp-perlmods installation I was expecting to have some file or directory containing the name mib2c No - the mib2c tool (and supporting directories) is part of the main 'net-snmp' RPM, not the 'net-snmp-perlmods' RPM. It *requires* the perl modules, but it isn't part of them. rpm -ql net-snmp-perlmods-5.2.1.2-1 Try rpm -ql net-snmp | grep mib2c Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: Get IP-addresses and names of hosts attached to Switchports
On Mon, 2005-10-17 at 13:15 -0400, Robert Story wrote: On Fri, 14 Oct 2005 12:14:39 +0200 Boris wrote: BF output: BF SNMPv2-SMI::mib-2.17.4.3.1.1.0.0.28.182.27.202 = Hex-STRING: 00 00 1C B6 BF 1B DF Any part of that look familiar? '28.182.27.202' looks like an ip address to me Close. It's actually 0:0:28:182:27:202, and is the Mac Address of the system concerned. This OID translates to BRIDGE-MIB::dot1dTpFdbAddress.0.0.28.182.27.202 and is the index of the Forwarding Database (dot1dTpFdbTable) BRIDGE-MIB::dot1dTpFdbPort (mib-2.17.4.3.1.2) will report which port traffic for that particular Mac Address will be sent to. You'd need to use another lookup to perform the Mac-IP mapping (e.g. the ipNetToMediaTable of some well-connected system, probably *not* the switch itself). Or use a local database. I don't know of any stanard MIBs that have ip/host mapping. That would be the DISMAN-NSLOOKUP-MIB. It's not particularly widely implemented, AFAIK. Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: timeout exponentially backoff ?
On Mon, 2005-10-17 at 12:34 -0400, Robert Story wrote: On Mon, 26 Sep 2005 13:32:41 +0100 Dave wrote: DS The backoff code was removed in 1998 Anyone have a good enough memory to remember why? After 7+ years? Are you nuts, Robert? I have enough difficulty remembering what got changed (and why) after 7 days! I don't have the time ATM to look up exactly when this change was made. (It took quite a bit of searching last time). But I seem to vaguely recall that it was probably part of Mike's SSI stuff. Seems like a good idea to me, eve it it's optional instead of the default. After so long with a fixed backoff, any change to the Net-SNMP timeout behaviour *must* be optional (and default to off). Yes - this is probably worth doing at some point. If you want to see it in 5.3, then you'd better get moving quickly! Otherwise it'll have wait until 5.4. Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: Help with exec and pass
Thanks Dave, LOL I liked your comments ;) I will make note of this as I am learning a lot about snmp with my little steps. -Regards Santanu -Original Message- From: Dave Shield [mailto:[EMAIL PROTECTED] Sent: Tuesday, 18 October, 2005 14:47 To: Santanu Misra Cc: net-snmp-users@lists.sourceforge.net Subject: RE: Help with exec and pass On Mon, 2005-10-17 at 10:38 +0200, Santanu Misra wrote: I am going to post the dummies how to on that. You are welcome to make your comments on that. Oh, you foolish, foolish person! You really shouldn't make suggestions like that :-) Here is the draft. 1. It should always reply with 3(three) sets of DATA - The OID - The Type (String or integer) A pass script can return more than just string or (signed) integer values. There are about eight or nine recognised types. You've missed the following: unsigned counter octet gauge objectid timetick ipaddress It doesn't support 64-bit counters (unfortunately), and you can probably safely ignore opaque. But the type returned must match the MIB definition. 2. It should not print anything or debug message if there is anything wrong it should just do an exit!! That's true for GET/GETNEXT requests. It doesn't hold for processing SET requests. 3. The script should be able to answer to -g and -n option only when that was passed otherwise do an exit. Nope - it should also handle -s, for SET requests. (I hope that you *have* read the man page!?) 4. When it is -g it should print the value for the OID it was asked. Example: # /usr/bin/lmstat.pl -g .1.3.6.1.4.1..1.0.0 .1.3.6.1.4.1..1.0.0 string QAC_runtime 5. When it is -n it should print the value for the next OID, # /usr/bin/lmstat.pl -n .1.3.6.1.4.1..1.0.0 .1.3.6.1.4.1..1.0.1 You'd probably be better off illustrating this with the next scalar-style OID - .1.3.6.1.4.1..1.1.0 It's unlikely that both .1.3.6.1.4.1..1.0.0 and .1.3.6.1.4.1..1.0.1 would be valid in any sensible MIB. (In fact, ending with two 0's is unusual, and you'd do better using ...1.1.0 and ...1.2.0 instead) integer 0 Note that with -n, the requested OID need not actually be valid for a -g request. The following invocations: /usr/bin/lmstat.pl -n .1.3.6.1.4.1..1.0.0 /usr/bin/lmstat.pl -n .1.3.6.1.4.1..1.0.0.0.0.0.0.0.0 /usr/bin/lmstat.pl -n .1.3.6.1.4.1..1.0.99 /usr/bin/lmstat.pl -n .1.3.6.1.4.1..1.1 should *all* return the same results (for .1.3.6.1.4.1..1.1.0) 6. When the last OID is reached it should exit. Or anything following. Dave - Visit our Internet site at http://www.reuters.com To find out more about Reuters Products and Services visit http://www.reuters.com/productinfo Any views expressed in this message are those of the individual sender, except where the sender specifically states them to be the views of Reuters Ltd. --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: How can I test perl modules?
On Tue, 2005-10-18 at 15:27 +0200, Roman Medina-Heigl Hernandez wrote: Radek Vokál wrote: I've tried to run sample perl module from http://net-snmp.sourceforge.net/tutorial/tutorial-5/toolkit/perl/perl_module.pl I've also modified the snmpd.conf to load the module # head /etc/snmp/snmpd.conf perl do /etc/snmp/perl_module.pl; snmpd doesn't display any error, so I guess it's loaded .. Do the following: kill/stop snmpd (if running as daemon), open a shell/xterm/etc and run snmpd in foreground: $ snmpd -f -Lsd -Lf /var/log/snmpd.log -p /var/run/snmpd -a Now you should see messages like: starting perl_module.pl perl_module.pl loaded ok That's exactly what I would love to see but still I don't get any line that the module is loaded [EMAIL PROTECTED] ~# snmpd -f -Lsd -Lf /var/log/snmpd.log -p /var/run/snmpd -a [EMAIL PROTECTED] ~# head /var/log/snmpd.log too many sensors. ignoring serial16 Turning on AgentX master support. NET-SNMP version 5.2.2.rc1 Connection from UDP: [127.0.0.1]:41819 Received SNMP packet(s) from UDP: [127.0.0.1]:41819 Received TERM or STOP signal... shutting down... In general your perl module can write to STDERR and it will be dumped into the former shell so it is a quick method for debugging. Note also that when you modify the module snmpd should be restarted, so you simply go to the shell, hit ctrl-c, and launch snmpd (in foreground) again. Regards, -Roman -- Radek Vokál [EMAIL PROTECTED] signature.asc Description: This is a digitally signed message part
Re: How can I test perl modules?
Radek Vokál wrote: at's exactly what I would love to see but still I don't get any line that the module is loaded [EMAIL PROTECTED] ~# snmpd -f -Lsd -Lf /var/log/snmpd.log -p /var/run/snmpd Run with -Dread_config -Dperl. +Thomas -- Thomas Anders (thomas.anders at blue-cable.de) --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: How can I test perl modules?
On Tue, 2005-10-18 at 15:27 +0200, Roman Medina-Heigl Hernandez wrote: Do the following: kill/stop snmpd (if running as daemon), open a shell/xterm/etc and run snmpd in foreground: $ snmpd -f -Lsd -Lf /var/log/snmpd.log -p /var/run/snmpd -a I'd actually suggest you run the agent using snmpd -f -Le rather than using syslog (-Lsd) and a log file (-Lf) The perl script you refer to outputs explicitly to STDERR, and while I haven't traced the output processing (or tried this in practise), I suspect it's probably safer to have the agent explicitly log to standard output too. Comparing your debug output to the perl control code, I'd expect the initializing perl message to be followed by either done initializing perl or an error: embedded perl support failed to initalize (see agent/snmp_perl.c:maybe_source_perl_startup()) Neither of these seems to occur. Strange What's in the file /usr/share/snmp/snmp_perl.pl ? Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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: mib2c
Hi Dave, I've patched net-snmp.spec at the top with %define _with_perl_modules 1 I've compiles everything, the net-snmp-perlmods-5.2.1.2-1.i386.rpm was generated, then I've installed it and it works now. Thanks so much, Eugen -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dave Shield Sent: Tuesday, October 18, 2005 4:39 AM To: Eugen Cc: 'Robert Story'; net-snmp-users@lists.sourceforge.net Subject: RE: mib2c On Mon, 2005-10-17 at 14:43 -0400, Eugen wrote: Thanks for your answer. Actually what I did was the following: - I've downloaded the package net-snmp-5.2.1.2-1.src.rpm from net-snmp web page I though you said you'd installed the RedHat-provided version of Net-SNMP ? - I've generated the binaries with rpm -bb net-snmp.spec The perl modules aren't built by default - you need to explicitly request that they be include (-with perl_modules). See the head of the net-snmp.spec file. So it seems that in 5.2.1.2 the net-snmp-utils is part of the other modules. Not quite - this is a difference between RedHat packaging and our arrangements, not the version of Net-SNMP. The RedHat RPMs distinguish between the setup required for just running the agent (net-snmp), and for using the command-line tools (net-snmp-utils). We don't - our net-snmp RPM covers both. When trying to run mib2c I received that error message than, following the instructions, I've installed the net-snmp-perlmods-5.2.1.2-1 directly from the binary downloaded on the same net-snmp web page (net-snmp-perlmods-5.2.1.2-1.i686.rpm since I didn't found the rpm source of this module). H. The problem with that particular RPM is that it doesn't specify which O/S it was built against. You may have better luck using one of the O/S-specific perlmods RPMs (rh9, fc2, fc4 or el4) Could this be caused by the fact that the binaries generated from the net-snmp source package have all the extension *i386.rpm versus net-snmp-perlmods-5.2.1.2-1.i686.rpm ? Sort of - it's more likely to be the O/S assumptions of that particular RPM don't match your system. That's why the other (i386) RPMs each specify which O/S they were built on. We'll need to ensure that the i686 version requirements are made clearer in the future. And another thing. After the net-snmp-perlmods installation I was expecting to have some file or directory containing the name mib2c No - the mib2c tool (and supporting directories) is part of the main 'net-snmp' RPM, not the 'net-snmp-perlmods' RPM. It *requires* the perl modules, but it isn't part of them. rpm -ql net-snmp-perlmods-5.2.1.2-1 Try rpm -ql net-snmp | grep mib2c Dave --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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 --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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
Binary that installs just the snmp utils
Hi: I want to install UCD-SNMP with just the SNMP utilities (primarily snmpget, snmpwalk) for Solaris 9. I donot want the agent or anything else. I have searched everywhere and not found one. 1. Does anybody know if such a binary exists 2. if not how do I create one. What do I need to do before I run configure? Thanks Ravi __ Yahoo! Mail - PC Magazine Editors' Choice 2005 http://mail.yahoo.com --- This SF.Net email is sponsored by: Power Architecture Resource Center: Free content, downloads, discussions, and more. http://solutions.newsforge.com/ibmarch.tmpl ___ 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
Multiple sets in one pdu
Hi, I'm facing a problem with multiple sets in one pdu, basically it's sequence issue. Say the pdu is set for 5 oids as following: oid_1.1 oid_2.1 oid_1.2 oid_2.2 oid_1.3 (oid_1 and oid_2 are variables in two different tables, .1/.2/.3 are indice) It seems the agent does some optimization, so instead of doing: set oid_1.1 set oid_2.1 set oid_1.2 set oid_2.2 set oid_1.3 it actually does following: set oid_1.1 set oid_1.2 set oid_1.3 set oid_2.1 set oid_2.2 This is not what I want, I want the original sequence. I'm using net-snmp 5.1. Has anyone seen this issue? I found in ChangeLog following, but seems it's not what I see. *5.0.1* Fixes: - the agent should properly handle requests in the right order. (e.g., restrict the agent to processing only one SET at a time)
Does anyone know what trap it is ?
hi ,all, ireceived some stange traps through trap receiver,the snmp trap 's source is 0.0.0.0,andtrap'soid is 1.3.6.1.4.8072.3.2.10,i lookup some references ,and get that this oid is locate in NET-SNMP.that is netSnmp(8072).netSnmpEnumerations(3).netSnmpAgentOIDs(2),and i wanna know that what the useness this trap it is and why it's source is 0.0.0.0,i will very appreciate anyone who can help me.and attachment is the pic of trap printscreen. attachment: snmp2.jpg