[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: how people implement <filter> in netconf over SOAP?




Zihong Lu wrote:

Thanks for replying.
Yes I am aware of that the filter selection model is vendor specific, as
described in the spec. Just wonder how you as a "vendor" implement this.
Here is how we implement it in YencaP:
The agent builds a super set of the config contained in the filter. This super set is the document that will be filtered.
This super set is a subset of the whole config (running, for example):
"filter config" smaller than "super set" smaller than "running"

How we build this super set is specific to our software. It uses a partial cache document and XPath to find the sub data models involved.

Then, a recursive algorithm is processing the filter, looking for content match nodes, selection nodes, ... at each XML level.
Each XML level matches a recursive level.

I know this is not optimal in terms of cpu consumption but it is easy to implement and it works (pretty fast).

Hope this helps,
Vincent

One way to do it would be to hard-code your specific selection model into
your SOAP skeleton code, for example, use type mapping in gSoap.  That will
avoid the XML parsing, but still, you are dealing with struct within struct
within struct, in C term.  The other way is what I have suspected, and
probably should be a new project like Ted Goddard suggested when I
communicated with him a while back, to have a toolkit that handles the
struct within struct within struct, to have a straight forward handling
function like

  getFilterData(target t)

The other problem that I have with the current netconf design is that it
doesn't allow vendors a chance to reinforce the schema, at the filter level
or at the config level, since there is no mechanism to show the data model
within the SOAP netconf wsdl, which leave a big hole in web services.  Not
sure how people deal with this problem.  Any suggesstion is highly
appreciated.

-Zihong



-----Original Message-----
From: Vincent Cridlig [mailto:vincent.cridlig@loria.fr]
Sent: Monday, January 30, 2006 12:04 AM
To: ??
Cc: zlu@extremenetworks.com
Subject: Re: how people implement <filter> in netconf over SOAP?


Yes, it is the vendor's job.

"filter" is totally independant of the underlying application protocol (SOAP, SSH, BEEP...). "filter" node selection model is Netconf-specific and, to my knowledge, no libraries are freely available.

Vincent


王罕 wrote:

I guess it should be the vedors' job.

ÔÚÄúµÄÀ´ÐÅÖÐÔø¾­Ìáµ½:


From: Zihong Lu <zlu@extremenetworks.com>
Reply-To: To: "'netconf@ops.ietf.org'" <netconf@ops.ietf.org>
Subject: how people implement <filter> in netconf over SOAP?
Date:Fri, 27 Jan 2006 11:07:12 -0800




I am looking into the implementation detail of netconf
over SOAP.  Wonder
how people implement <filter> in SOAP. Can any one please
shed some
light?

In general, SOAP is a service function based protocol. It
should be
possible to be message-based, but it will not be as easy
to implement, and
the SOAP benefit will mostly be lost. For example, the
<get-config> part,
we have a message looks like this:

<soapenc:Body>
<rpc>
    <get-config>
         <fource/>
             <filter>
                 <users/>
             </filter>

--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>



begin:vcard
fn:Vincent Cridlig
n:Cridlig;Vincent
org:LORIA - INRIA Lorraine;Madynes
adr:;;;Nancy;;;France
email;internet:cridligv@loria.fr
title:PhD Student
tel;work:+33 (0)3 83 59 20 48
url:http://www.loria.fr/~cridligv
version:2.1
end:vcard