I assume you mean "smidump -f xsd" on the rfc2863 mib (after running it through smistrip) and not really "smilint -f xsd". I did this and the parses generated an error:
/IF-MIB:1134: index element `ifRcvAddressAddress' of row `ifRcvAddressEntry' must have a size restriction
So I added a size restriction to the mib and ran smidump -f xsd again and there were no errors. But, when I opened the xsd, there was no "ifTable" or "IfEntry", just the "ifEntry". If fact, the xsd output never seems to include tables, just the entries. Is this intentional? I am using libsmi from cvs.
Steve
-----Original Message----- From: Frank Strauss [mailto:strauss@ibr.cs.tu-bs.de] Sent: Wednesday, October 23, 2002 11:47 AM To: Steve Bucknam Cc: libsmi@ibr.cs.tu-bs.de Subject: Re: [libsmi] xsd output
Hi!
Steve> I am generating the xsd for the version of RCF1213 retrieved from
Steve> mibcentral. I get no errors or warnings, yet "ifTable" has no Steve> entry. The "ifEntry" is present and contains the sequence of OIDs
Steve> that should appear under "IfEntry". "IfEntry" is also missing.
Steve> Is this a design feature to minimize unnecessary nodes, or a bug?
Well, RFC 1213 is a bug. :-)
We don't care a lot about obsolete SMIv1 MIBs. The ifTable is defined in the current Standards Track version of the IF-MIB which is published in RFC 2863.
If you use the IF-MIB (it also comes with libsmi), then `smilint -f xsd' should produce a better result.
-frank
Hi!
Steve> I assume you mean "smidump -f xsd" on the rfc2863 mib (after running it Steve> through smistrip) and not really "smilint -f xsd".
Sorry for the confusion. Yes, I meant smidump. Note, that you don't have to strip MIBs from the RFC documents. All IETF Standard MIBs come with libsmi. So, if libsmi is installed, you can simply use `smidump -f xsd IF-MIB'.
Steve> I did this and the parses generated an error:
Steve> /IF-MIB:1134: index element `ifRcvAddressAddress' of row Steve> `ifRcvAddressEntry' must have a size restriction
Yes, there is a (minor) SMIv2 violation in the IF-MIB.
Steve> So I added a size restriction to the mib and ran smidump -f xsd again Steve> and there were no errors.
I'd suggest *not* to modify existing MIBs (this way). Usually supporting an existing (probably slightly invalid) Standard MIB is more important than supporting a correct `Non-Standard' MIB.
Steve> But, when I opened the xsd, there was no "ifTable" or Steve> "IfEntry", just the "ifEntry". If fact, the xsd output never Steve> seems to include tables, just the entries. Is this Steve> intentional?
Yes, it is intentional. We felt no need for the additional nesting of explicit table elements. E.g., addressing (the elements of) a table in XPath would be no easier (IF-MIB/ifTable/ifEntry vs. IF-MIB/ifEntry).
-frank
participants (2)
-
Frank Strauss -
Steve Bucknam