[2002/11/14 02:47:21] #net-snmp <hampa> I created a table with only one index in like this : ProfileEntry ::= SEQUENCE { row OCTET STRING }
[2002/11/14 02:47:35] #net-snmp <hampa> when I walk it I get strange output
[2002/11/14 02:47:42] #net-snmp <hampa> profileTable.profileEntry.row."." "zero zero zer"
[2002/11/14 02:47:47] #net-snmp <hampa> the "." part
[2002/11/14 02:47:55] #net-snmp <hampa> I would like a 1 there instead
[2002/11/14 03:10:31] #net-snmp <jbpn> Errr, right.
[2002/11/14 03:10:46] #net-snmp <jbpn> Do you mean that you want the row indexed by an integer then?
[2002/11/14 03:29:38] #net-snmp <jbpn> Or do you mean you want the string "1"?
[2002/11/14 03:54:25] #net-snmp <hampa> I would like row.1 "asdfasdf"
[2002/11/14 03:55:32] #net-snmp <hampa> so.. indexed by an integer, like all my other tables that have more that one values in there SEQUENCE
[2002/11/14 04:08:08] #net-snmp <jbpn> Okay, well, you need to have two entries in the SEQUENCE then.
[2002/11/14 04:08:38] #net-snmp <jbpn> An integer index (which is not-accessible) and the string column
[2002/11/14 04:11:20] #net-snmp <hampa> is not-accessible something I define in the MIB?
[2002/11/14 04:31:51] #net-snmp <jbpn> Yes, it's the appropriate value of the MAX-ACCESS clause for an index
[2002/11/14 04:32:01] #net-snmp <jbpn> (usually)
[2002/11/14 04:57:15] #net-snmp <hampa> does the xxxId in a SEQUENCE have to be at xxxEntry 1
--> Fleury has joined #net-snmp
[2002/11/14 05:01:49] #net-snmp <Fleury> hi..
[2002/11/14 05:07:36] #net-snmp <hampa> it worked with Id on a different number.. not sure if its the "correct" way of doing it
[2002/11/14 05:39:31] #net-snmp <rstory> hampa: xxxEntry must be 1.
[2002/11/14 05:47:08] #net-snmp <hampa> I mean if I can have xxxEntry :: = SEQUENCE { row OCTET STRING, xxxId INTEGER }
[2002/11/14 05:47:14] #net-snmp <hampa> so xxxId is 2
[2002/11/14 05:48:18] #net-snmp <rstory> ah... ok
--> Fleury has joined #net-snmp
[2002/11/14 05:53:33] #net-snmp <jbpn> you can have it whichever way you like. What matters is the INDEX clause
[2002/11/14 05:54:03] #net-snmp <jbpn> But conventionally, the indexes of a table are its first columns.
--> wes has joined #net-snmp
[2002/11/14 05:56:06] #net-snmp <Fleury> hi.. I´m having some problens in C++ object, create with the code of snmpwalk... I´m using UCD 4.2.6, SunOS 5.8.. The error occur in function init_snmp("snmpapp");, inside of snmp_parse_args.... Anyone know what can cause this? Compiler problem?
[2002/11/14 05:56:55] #net-snmp <Fleury> This error occur only when i create the object.
[2002/11/14 05:57:53] #net-snmp <wes> I'd have to see the error.
[2002/11/14 05:59:39] #net-snmp <jbpn> Can you give us the stack backtrace from gdb?
[2002/11/14 06:00:53] #net-snmp <Fleury> ok...
[2002/11/14 06:02:31] #net-snmp <Fleury> Wes and jbpn, in query.
[2002/11/14 06:02:48] #net-snmp <jbpn> ??
[2002/11/14 06:03:19] #net-snmp <Fleury> #0 0xfef3a448 in nvmatch () from /usr/lib/libc.so.1
[2002/11/14 06:03:19] #net-snmp <Fleury> #1 0xfef3a4fc in getenv () from /usr/lib/libc.so.1
[2002/11/14 06:03:19] #net-snmp <Fleury> #2 0xfef716c4 in locale_per_category () from /usr/lib/libc.so.1
[2002/11/14 06:03:19] #net-snmp <Fleury> #3 0xfef6efcc in _real_setlocale () from /usr/lib/libc.so.1
[2002/11/14 06:03:19] #net-snmp <Fleury> #4 0xfef6ecac in setlocale () from /usr/lib/libc.so.1
[2002/11/14 06:03:19] #net-snmp <Fleury> #5 0xff331008 in init_snmp (type=0xff362e30 "snmpapp") at snmp_api.c:648
[2002/11/14 06:03:21] #net-snmp <Fleury> #6 0xff34a920 in snmp_parse_args (argc=5, argv=0xffbef988,
[2002/11/14 06:03:24] #net-snmp <Fleury> session=0xffbefaa0, localOpts=0x1400
,
[2002/11/14 06:03:25] #net-snmp <Fleury> proc=0x12be4 ) at snmp_parse_args.c:416
[2002/11/14 06:03:27] #net-snmp <Fleury> #7 0x00012ea0 in OpenWalk__8snmpwalki (this=0xffbefa9c, argc=5)
[2002/11/14 06:03:29] #net-snmp <Fleury> at snmpwalk.c:105
[2002/11/14 06:03:31] #net-snmp <Fleury> #8 0x00012990 in main () at testeSNMP1.cpp:8
[2002/11/14 06:03:40] #net-snmp <wes> Um, wait a minute.
[2002/11/14 06:03:55] #net-snmp <wes> init_snmp() never calls snmp_parse_args, so have you modified the source?
[2002/11/14 06:05:15] #net-snmp <Fleury> snmp_parse_args calls init_snmp.....
[2002/11/14 06:05:32] #net-snmp <wes> Uh, sorry. read it in the wrong order.
[2002/11/14 06:05:42] #net-snmp <wes> what command line arguments are yougiving to snmp_parse_args?
[2002/11/14 06:06:58] #net-snmp <Fleury> the same in the snmpwalk -> arg = snmp_parse_args(argc, argv, &session, "C:", optProc)
[2002/11/14 06:08:49] #net-snmp <wes> I'm not sure. Your stack trace shows it dying in setlocale, but I'm not sure why it would. It's likely that stack trace is actually after the real bug was hit.
[2002/11/14 06:08:58] #net-snmp <wes> you might link with -lefence and see if you get a different result.
[2002/11/14 06:09:08] #net-snmp <Fleury> ok.
[2002/11/14 06:09:49] #net-snmp <jbpn> why is localOpts shown as pointing at an address that is OOB?
[2002/11/14 06:10:03] #net-snmp <jbpn> that doesn't sound right
[2002/11/14 06:10:32] #net-snmp <wes> I was wondering that too, but the function call looks right.
[2002/11/14 06:10:37] #net-snmp <jbpn> Mmm.
[2002/11/14 06:10:42] #net-snmp <jbpn> Strange though.
[2002/11/14 06:19:52] #net-snmp <Fleury> Don´t have the lefence library.
[2002/11/14 06:20:07] #net-snmp <wes> its free if you want to find it.
[2002/11/14 06:20:40] #net-snmp <Fleury> ok.. i will try.
--> Yruelf has joined #net-snmp
[2002/11/14 06:43:15] #net-snmp <Fleury> Problens with compile efence... ld: elf error: file libefence.a: elf_begin: Format error: archive fmag.... How can I solve this?
[2002/11/14 06:49:12] #net-snmp <wes> err.
[2002/11/14 06:49:14] #net-snmp <wes> no idea.
[2002/11/14 06:49:17] #net-snmp <wes> give up?
[2002/11/14 06:49:28] #net-snmp <wes> use dmalloc (harder to use) or something else instead?
[2002/11/14 06:49:36] #net-snmp <wes> track down the bug through painful checking?
[2002/11/14 06:49:40] #net-snmp <wes> don't know.
[2002/11/14 06:54:01] #net-snmp <jbpn> I think the best approach would be to look at _exactly_ how your application differs from snmpwalk.
[2002/11/14 06:54:06] #net-snmp <jbpn> Maybe write it in C first.
[2002/11/14 06:58:02] #net-snmp <Fleury> In C, its work.
[2002/11/14 07:00:18] #net-snmp <jbpn> Your optProc -- does that have extern "C" linkage?
[2002/11/14 07:00:25] #net-snmp <jbpn> It needs to.
[2002/11/14 07:01:02] #net-snmp <Fleury> Another thing... where i found a net-snmp distribution specific for SunOS or Solaris???
[2002/11/14 07:01:22] #net-snmp <jbpn> There isn't a distribution specific to SunOS/Solaris.
[2002/11/14 07:01:39] #net-snmp <jbpn> There may be some pre-built binaries but it's all the same code.
[2002/11/14 07:01:48] #net-snmp <jbpn> (Admittedly peppered with conditionals)
[2002/11/14 07:01:48] #net-snmp <Fleury> I test the code in Windows.. it works...Strange.
[2002/11/14 07:02:44] #net-snmp <Fleury> ok..
[2002/11/14 07:03:37] #net-snmp <nostaw> ACTION is back (gone 16:22:59)
[2002/11/14 07:06:51] #net-snmp <Fleury> thanks.. see you next week. Bye...
--> wes has joined #net-snmp
--> miike has joined #net-snmp
--> wdebruij has joined #net-snmp
[2002/11/14 10:04:04] #net-snmp <wdebruij> Hi all
[2002/11/14 10:04:24] #net-snmp <wdebruij> I've got a quick question for the crowd...
[2002/11/14 10:05:04] #net-snmp <wdebruij> I need to be able to SET some useful variables in a couple of days. It's not of vital importance which ones, as long as they actually manipulate the system
[2002/11/14 10:05:36] #net-snmp <wdebruij> does anyone know how to do this? apart from sysName.0 all variables in the standard net-snmp package are set read only
[2002/11/14 10:06:08] #net-snmp <wes> many are read-write, but many of the standard MIB-II set are still read-only.
[2002/11/14 10:06:18] #net-snmp <wes> (sysLocation and sysContact are)
[2002/11/14 10:06:27] #net-snmp <wes> many USM and VACM mibs are wirtable, for instance.
[2002/11/14 10:06:59] #net-snmp <wdebruij> doing a quick grep in the mibs dir didn't give much hope
[2002/11/14 10:07:30] #net-snmp <wdebruij> do I have to explicitly add these MIBs to the snmpd list?
[2002/11/14 10:08:25] #net-snmp <wdebruij> oh and the route info and networking MIBS (if, ip, tcp, etc) don't have r/w variables, right?
[2002/11/14 10:08:39] #net-snmp <wes> right.
[2002/11/14 10:08:53] #net-snmp <wdebruij> hmm too bad
[2002/11/14 10:08:54] #net-snmp <wes> many don't have reda-write variables, as defined by the standards. The ones that do we often haven't implemented yet.
[2002/11/14 10:09:29] #net-snmp <wdebruij> many of these variables need to be SETtable on pure hardware devices, such as switches and routers.
[2002/11/14 10:10:16] #net-snmp <wdebruij> and I'm trying to mimic a routed infrastructure using standard hosts
[2002/11/14 10:10:19] #net-snmp <wdebruij> so I
[2002/11/14 10:10:29] #net-snmp <wdebruij> 'd like to set some variables according to input
[2002/11/14 10:11:53] #net-snmp <wdebruij> Is there any variable that you think I might be able to make read/write within a couple of hours?
[2002/11/14 10:12:04] #net-snmp <wdebruij> I'll upload a patch if I get it working ofcourse
[2002/11/14 10:12:45] #net-snmp <wdebruij> preferable something not SNMP related (so no USM or VACM)
[2002/11/14 10:18:32] #net-snmp <wdebruij> no?
[2002/11/14 10:19:07] #net-snmp <wes> Um, maybe.
[2002/11/14 10:19:17] #net-snmp <wes> lots of things should be easily writable it just hasn't been done.
[2002/11/14 10:19:25] #net-snmp <wes> it's actually something I've wanted to do.
[2002/11/14 10:19:38] #net-snmp <wdebruij> ok
[2002/11/14 10:19:52] #net-snmp <wes> you could try the routing table.
[2002/11/14 10:19:56] #net-snmp <wes> it'd be a good start.
[2002/11/14 10:20:07] #net-snmp <wes> It's creation that is hard. Modification is easy, generally.
[2002/11/14 10:20:14] #net-snmp <wdebruij> I will then, thanks. Just one more thing
[2002/11/14 10:20:49] #net-snmp <wdebruij> the MIB text files (the ones that say read-only) are the real standards in ASN.1 right
[2002/11/14 10:21:02] #net-snmp <wdebruij> or are they only there as reference to the current implementation?
[2002/11/14 10:23:08] #net-snmp <miike> Not exactly sure of the question but, the MIB text files are the standard definitions (IETF RFC's), the implementations should try to match them as much as possible
[2002/11/14 10:24:18] #net-snmp <wdebruij> so making something read/write that is defined as read in the standards isn't a good thing. I justed wanted to make sure that everything defined as read-only should stay read-only
[2002/11/14 10:25:03] #net-snmp <miike> Generally speaking, it should match the MIB text files.
[2002/11/14 10:25:26] #net-snmp <wdebruij> I'll try to create one or more writable variables tomorrow. I just wanted to be sure that it is doable in a short timespan (at least in theory).
[2002/11/14 10:33:58] #net-snmp <wdebruij> thanks for your feedback, I'll let you know if I have some working code
[2002/11/14 10:34:54] #net-snmp <wes> actually, I tihnk the route table might be writale now.
[2002/11/14 10:34:56] #net-snmp <wes> try it.
[2002/11/14 10:35:05] #net-snmp <wes> (mibII/ip.c seems to indicate iti s)
[2002/11/14 10:35:21] #net-snmp <wdebruij> oh, that would be great. I'll try it
[2002/11/14 10:37:32] #net-snmp <wdebruij> I'm looking at the 5.0.3 code and can only see WIN32 write functions for forwarding and default TTL
[2002/11/14 10:38:21] #net-snmp <wes> well, I see RWRITE declared vars outside the WIN32 list. I think the write functions are in a differnt file (route_write.c)
[2002/11/14 11:03:51] #net-snmp <wdebruij> It didn't seem to work from Linux. I'll delve a bit deeper then
[2002/11/14 11:03:57] #net-snmp <wdebruij> For now, thanks again. Bye
--> miike has joined #net-snmp
--> linfel has joined #net-snmp
--> ian_c has joined #net-snmp
[2002/11/14 15:02:09] #net-snmp <ian_c> is anyone here?
--> kerebri has joined #net-snmp
--> rstory has joined #net-snmp
--> jbpn has joined #net-snmp