hinkey / libnetconf Goto Github PK

What steps will reproduce the problem?
1. Go to 
http://libnetconf.googlecode.com/git/doc/doxygen/html/d9/d87/install.html
2. The last option for the configure script "--with-examples" is obsolete now 
as the example client and server have been moved to netopeer project

What steps will reproduce the problem?
1. Linux system does not have xsltproc installed.
2. $ lnctool -h
3. lnctool exits with: "Missing pyang tool, please install it."

What is the expected output? What do you see instead?
Since pyang _is_ installed, I expect to see:
   "Missing xsltproc tool, please install it."
Instead, I see:
   "Missing pyang tool, please install it."

What version of the product are you using? On what operating system?
   I cloned at commit: 4ae9f8bcc37d92f10dc27bb5690bb6b3688ed5b1

   OS= Linux 3.8.0-25-generic #37-Ubuntu x86_64 x86_64 x86_64 GNU/Linux

Please provide any additional information below.

What steps will reproduce the problem?
1. running.xml is your running configuration
2. connect to server
3. send edit-config.xml request

What is the expected output? What do you see instead?
In resulting running configuration "id" element is missing in container 
"exporter", "id" is key element for "exporter".
Container "exporter" should have child "id".

Please use labels and text to provide additional information.

When xc:operation="merge" attribute is removed or moved to "exporter" element 
the behavior is correct.

What steps will reproduce the problem?
1. Start the client with a empty running datastore and the attached module
2. Send the first message of the "test" attachment (creating 3 lists and 1 
leaf).
3. Send the second message of the "test" attachment (replacing the previous 
values).

What is the expected output? What do you see instead?
For the first message, I expected callbacks for adding the four nodes (lists 
and leaf). For the second message, I expected callbacks for changing the leaf, 
deleting the old lists and adding the new lists. But I obtained a weird 
behaviour for the second message (described in test attachment).

What version of the product are you using? On what operating system?
I am using the branch master commit 3f33930cd3a8 (Thu Sep 5 13:50:40 2013 
+0200) on a Gentoo Linux.

What steps will reproduce the problem?
1.Add a new module in netopeer-sl, it defines a RPC with a list as input. eg: 
{ rpc get-port-stats {
     input {
      list port {
        .....
      }
      ....
     }
}

2. Use lnctool to generate transapi and implement the RPC, then build and 
install the module, restart the sshd server.   

3. use netopeer-cli to test the rpc, use user-rpc for test, message is
<get-port-stats xmlns="mynamespace">
<port>something</port>
<port>something otehr</port>
<port>something else</port>
</get-port-stats>

What is the expected output? What do you see instead?
I want to see all the port statistics in my input list, however
i only get the statistics of the last port in the list. 

What version of the product are you using? On what operating system?
some earlier version, Linux.  

Please provide any additional information below.
I checked the root cause, i found codes like this(libnetconf/src/datastore.c, 
line 6111) 
nc_reply* ncds_apply_rpc(ncds_id id, const struct nc_session* session, const 
nc_rpc* rpc)
{
...
   case NC_OP_UNKNOWN:
       ....
       /* copy xmlnode to the input arg list */
   ....
...
}

if the xml nodes are in a node list, the code above will repeat several times 
and the arg[] only stores the last one. 
In my example, only "<port>something else</port>" is  passed to my module, then 
i only got the last port's statistics. 

In my opinion, we should use xmlCopyNodeList instead of xmlCopyNode, and we 
should set the right step while copying, or may occur unexpected error. 

What steps will reproduce the problem?
1. Start the server and client
2. Send a edit-config operation with three configurations (that will call three 
callbacks) where the second callback return EXIT_FAILURE.

What is the expected output? What do you see instead?
I expected the third callback not be called (due the failure return of the 
second callback) and the configuration set on database consisting only of the 
first configuration.
I see the third callback is not called (ok), but the database set ALL the 
configurations (got sending a get-config after the edit-config described above)!
This behaviour is very bad because:
1. If the device error condition was gone and I try apply the second 
configuration again, I will can't! (because will be no differences between  
this configuration and configuration on datastore, therefore the callback will 
not be called)
2. The third configuration will never be applied because the callback was not 
called in the first time and in the following times there will not diffs 
(analogous to the case 1). And the user will not know that configuration was 
not applied!

What version of the product are you using? On what operating system?
I am using the branch master commit 3f33930cd3a8 (Thu Sep 5 13:50:40 2013 
+0200) on a Gentoo Linux.

What steps will reproduce the problem?
1. ./configure --enable-shared --prefix=/ws/shhv-bgl/libnetconf/ 
CURL_CFLAGS=/ws/shhv-bgl/curl/bin/ CURL_LIBS=/ws/shhv-bgl/curl/lib/ 
--with-libxml2=/ws/shhv-bgl/libxml2/bin 
--with-libxslt=/ws/shhv-bgl/libxslt/bin/ --with-libssh2=/ws/shhv-bgl/libssh2/
2. make


What is the expected output? What do you see instead?
What i get : 
make
./libtool --tag=CC --quiet --mode=compile gcc -Wall 
-I/ws/shhv-bgl/libssh2//include -I/ws/shhv-bgl/libxslt/include 
-I/ws/shhv-bgl/libxml2//include/libxml2 -I/ws/shhv-bgl/libxml2//include/libxml2 
 -O3  -DNC_WORKINGDIR_PATH=\"//var/lib/libnetconf/\" -DSETBIT=0  
-DRCSID=\""built from git 13013a72a3e64b0c82b8"\" 
-I/ws/shhv-bgl/HNR/python/libnetconf/src  -fPIC -c src/error.c -o 
.obj/src/error.lo
In file included from src/error.c:50:
src/netconf_internal.h:303: error: syntax error before "pthread_rwlock_t"
src/netconf_internal.h:303: warning: no semicolon at end of struct or union
src/netconf_internal.h:306: error: syntax error before '}' token
make: *** [.obj/src/error.lo] Error 1

What version of the product are you using? On what operating system?
0.0.6

Please provide any additional information below.

What steps will reproduce the problem?
1.Connect with netopeer-cli
2.Enter username & passoword
3.

What is the expected output? What do you see instead?

Expected output is be able to establish session. I see a malformed error message


What version of the product are you using? On what operating system?

Latest version


Please provide any additional information below.

Feb 28 11:30:08 reinaldo-sony ncserver[17669]: E| Datastore preparing failed.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: D| Shared memory key: -4
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: D| Shared memory ID: 131074
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-inet-types.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-yang-types.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-netconf-monitoring.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-netconf-notifications.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model nc-notifications.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model notifications.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-netconf-with-defaults.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| ncds_features_parse: no 
feature definitions found in data model ietf-netconf-acm.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: D| ncds_set_validation: Relax NG 
validator set (//var/lib/libnetconf//ietf-netconf-acm-data.rng)
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: D| ncds_set_validation: 
Schematron validator set (//var/lib/libnetconf//ietf-netconf-acm-schematron.xsl)
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: V| Checking the default Events 
stream path //var/lib/libnetconf//streams/.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: E| Wrong transAPI version of the 
module //var/lib/libnetconf/server//toaster.so. Have 3, but 4 is required.
Feb 28 11:34:50 reinaldo-sony ncserver[21472]: E| Datastore preparing failed.

What steps will reproduce the problem?
1. Connect to server.
2. Send copy-config message:
<rpc xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="12">
  <copy-config>
        <source>
                <invalid/>
        </source>
        <target>
                <running/>
        </target>
</copy-config>
</rpc>


What is the expected output?
<rpc-reply xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="13">
  <rpc-error>
    <error-tag>bad-element</error-tag>
    <error-type>application</error-type>
    <error-severity>error</error-severity>
    <error-message>An element value is not correct.</error-message>
    <error-info>
      <element>source</element>
    </error-info>
  </rpc-error>
</rpc-reply>

What do you see instead?
<rpc-reply xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="12">
  <ok/>
</rpc-reply>


Please use labels and text to provide additional information.

What steps will reproduce the problem?

netconf> edit-config

  Select target datastore (running|startup|candidate|url:<dsturl>): running

  Type the edit configuration data (close editor by Ctrl-D):
<toaster xmlns="http://netconfcentral.org/ns/toaster"/>

  Result OK
netconf> user-rpc

  Type the content of a RPC operation (close editor by Ctrl-D):
<?xml version="1.0"?>
<make-toast xmlns="http://netconfcentral.org/ns/toaster" />

NETCONF error: resource-denied (protocol) - Request could not be completed 
because of insufficient resources

Feature description:

Optimize filtering on <get> (and others) operation. Currently, data are 
retrieved from all datastore parts even if they are then completely filtered 
out.

Hi guys,

Ran into this issue while testing a couple of configurations.

** What steps will reproduce the problem?
1. Send an invalid (syntax incorrect) XML configuration file

** What is the expected output? What do you see instead?
Expected an error message ('Malformed message') to be returned to the client.
Instead, the session gets dropped without any error message on the client side. 


** What version of the product are you using? On what operating system?
Using the tip of libnetconf master branch at this moment (commit tag 
e559385d7f33052df5a4d16adbb65d679f7111aa)

** Please provide any additional information below.
From within the code (libnetconf/src/session.c) at function 
'nc_session_receive' the malformed_msg label is hit as expected. However the 
call to 'nc_session_send_reply' has the 'rpc' parameter set to NULL which 
immediately fails without sending any error message to the client.

Regards
Dirk vd Merwe

Using empty.yang data model (attached) and empty data store.

get-config running without with-default set (as well as set to explicit and 
trim) returns empty data.

get-config with with-default set to either of report-all or report-all-tagged 
returns empty container <np/> in data.

Branch master commit 2076c60b5fa5

Feature description:

Implement functions to support Call Home mechanism (Reverse SSH for NETCONF 
[1]) for both client and server.

[1] http://tools.ietf.org/html/draft-ietf-netconf-reverse-ssh-03

Feature description:

Check ncds_apply_rpc2all() and all called functions and avoid problems caused 
by a concurrent usage of this function by multiple threads.

Feature description:

libnetconf curretnly uses System V shared memory. Despite this is the most 
common type of shared memory mechanism, configure should detect that it is 
available and there should be possibility to switch to the POSIX shared memory. 
Consider a complete switch to the POSIX shared memory.

In general, utilization of the shared memory in libnetconf should be reviewed 
to allow using libnetconf in multiple independent applications running on the 
system. Currently, all applications using libnetconf are always connected.

Feature description:

Get rid of libssh2_poll() - deprecated function is needed to poll data from 
multiple SSH channels, but it is marked as deprecated and it will be removed in 
future version of libssh2. It also comes with some problems like inconsistency 
of results compared to poll() (caller's poll() says "we have a data" and 
libssh2_poll() says "no, there are no data to read").

Use POSIX compatible variants of function instead of asprinf and vasprintf.

What steps will reproduce the problem?
1. Create data model and datastore with filled startup configuration datastore.
2. Start server (incoming connection).
3. Send get-config rpc message with running as a source.

What is the expected output?

Same data as in startup datastore.

What do you see instead?

Empty string.

Please use labels and text to provide additional information.

There were no libxml2 CFLAGS in lncdatastore/Makefile.

Created and pushed patch into the origin/compilation-bugfix branch that should 
solve this problem.

TODO: configure script was NOT regenerated, the configure.in file was modified 
and commited only... Please commit Your current output of autoconf to avoid 
version change.

A larger running configuration returned at transapi_init of a transapi module 
is leading to a segfault in fill_default (src/with_defaults.c:612). Attached is 
the according valgrind.log.

During the recursion the variable created_local is reallocated. If the 
reallocation happens during the recursion, the pointer gets invalid.

What steps will reproduce the problem?
1. apply attached patch (0001-modified... to reduce initially allocated memory)
2. run netconf server

Proposed solution:
0001-pointer-made-static...patch makes the pointer static, thus reallocation 
works fine.

What steps will reproduce the problem?
1. Apply edit-config message with test-option=test-then-set to module without 
generated schemes.


What is the expected output?
rpc message with OK or rpc-error.

What do you see instead?
NCDS_RPC_NOT_APPLICABLE



Please use labels and text to provide additional information.

What steps will reproduce the problem?
1. Issue a get-schema rpc. 
<rpc message-id="m-0" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0">
<get-schema xmlns="urn:ietf:params:xml:ns:yang:ietf-netconf-monitoring">
<format>yang</format>
<identifier>ietf-netconf-acm</identifier>
<version>2012-02-22</version>
</get-schema>
</rpc>
2. Wait for result rpc


What is the expected output? What do you see instead?

Expected response is rpc-reply conforming to the ietf-netconf-monitoring, with 
namespace urn:ietf:params:xml:ns:yang:ietf-netconf-monitoring.

Incorrect returned value was:

<?xml version="1.0" encoding="UTF-8"?>
<rpc-reply xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="m-0">
  <data xmlns="urn:ietf:params:xml:ns:netconf:base:1.0">module "ietf-netconf-acm" {
  namespace "urn:ietf:params:xml:ns:yang:ietf-netconf-acm";
  prefix "nacm";
  ...

Which contains a data element with namespace 
urn:ietf:params:xml:ns:netconf:base:1.0.
The correct namespace should be 
urn:ietf:params:xml:ns:yang:ietf-netconf-monitoring.



Build on master

Please provide any additional information below.

What steps will reproduce the problem?
1.
Yang File :
-----------
    container college-info {

        leaf college-name {         
            must "current() = 'NIT'";
            type string;
        }
    }
2.
Valid Configuration Data :
--------------------------
netconf> edit-config running
<college-info xmlns="http://www.adtran.com/ns/college"  
xmlns:xc="urn:ietf:params:xml:ns:netconf:base:1.0" xc:operation="create">
<college-name>NIT</college-name>
</college-info>

3.
Invalid Configuration Data :
----------------------------
netconf> edit-config running
<college-info xmlns="http://www.adtran.com/ns/college"  
xmlns:xc="urn:ietf:params:xml:ns:netconf:base:1.0" xc:operation="create">
<college-name>ABC</college-name>
</college-info>

4.
What is the expected output? What do you see instead?
Expected output :
-----------------
edit-config should pass for the valid configuration data
And
edit-config should fail for the invalid configuration data

Actual output :
----------------
edit-config is pass with even valid and invalid configuration data.

5. 
What version of the product are you using? On what operating system?
Version :
---------
libnetconf 0.8.0

Operating System :
-------------------
Ubuntu 12.04.4 LTS

6.
Please provide any additional information below.
Root Cause Analysis :
--------------------
Not properly parsing the result of xsltApplyStylesheet in datastore.c file, 
validate_ds function.

Code :
------
The below code/line is causing the problem.

if ((result = xmlXPathEvalExpression(BAD_CAST 
"/svrl:schematron-output/svrl:successful-report/svrl:text", ctxt)) != NULL) {

7.
Possible Solution:
------------------
By changing the above code as below, it is working properly.

if ((result = xmlXPathEvalExpression(BAD_CAST 
"/svrl:schematron-output/svrl:failed-assert/svrl:text", ctxt)) != NULL) {

8.
Summary :
Replaced "successful-report" with "failed-assert".

What steps will reproduce the problem?
1. NC client has: "include <libnetconf.h>"
2. Embedded target is cross-compiled with "-Werror -Wstrict-prototypes".

What is the expected output? What do you see instead?
Expected build to produce no warnings.
Instead:
   "error: function declaration isn't a prototype
    [-Werror=strict-prototypes]
    cc1: all warnings being treated as errors"

What version of the product are you using? On what operating system?
VERSION 0.5.99
commit f87d9fb67c113a610297d028ddc07df6c41bc3d1
Author: Radek Krejci <[email protected]>
Date:   Fri Aug 16 09:51:10 2013 +0200

OS
Linux 3.0.17 #3 SMP Tue May 1 13:58:54 PDT 2012 i686 i686 i386 GNU/Linux

Please provide any additional information below.

What steps will reproduce the problem?
1. ./configure && make rpm
Output:
make[1]: Entering directory 
`/home/local/cejkat/pokus/RPMBUILD/BUILD/libnetconf-0.5.99'
./libtool --tag=CC --quiet --mode=compile gcc -Wall -I/usr/include/libxml2 
-I/usr/include/libxml2  -O3  -I/usr/include/dbus-1.0 
-I/usr/lib64/dbus-1.0/include   -DNC_WORKINGDIR_PATH=\"//var/lib/libnetconf/\" 
-DSETBIT=0  -DRCSID=\""built from git e268d5694d29fc0f6e6d"\" 
-I/home/local/cejkat/pokus/RPMBUILD/BUILD/libnetconf-0.5.99/src  -fPIC -c 
src/callbacks.c -o .obj/src/callbacks.lo
libtool: Version mismatch error.  This is libtool 2.4.2, but the
libtool: definition of this LT_INIT comes from libtool 2.2.6b.
libtool: You should recreate aclocal.m4 with macros from libtool 2.4.2
libtool: and run autoconf again.
make[1]: *** [.obj/src/callbacks.lo] Error 63

What is the expected output? What do you see instead?
Created RPM package is expected.

Please use labels and text to provide additional information.

The problem is caused by old version of autotools on development PC.
Attached patch solved the problem for two version of autotools:
GNU libtool 2.2.6b
GNU Autoconf 2.63
and
GNU libtool 2.4.2
GNU Autoconf 2.68

The library libreadline is not checked during configure. It is as well not in 
the documentation as dependency.

What steps will reproduce the problem?
1.Download and extract libnetconf contents to a folder
2. cd libnetconf; ./configure


What is the expected output? What do you see instead?
Expected: 
Configured successfully.

What i see:
checking for xslt-config... yes
checking for libssh2_version in -lssh2... no
configure: error: Missing libssh2.

Contents of config.log:
configure:11635: checking for libssh2_version in -lssh2
359 configure:11660: gcc -o conftest 
-I/usr/cisco/packages/libxslt/libxslt-1.1.6/include 
-I/usr/cisco/packages/libxml2/current/include/libxml2 -I/usr/cisco/packag    
es/libxml2/current/include/libxml2  -O3  -DRCSID=\"$(IDGIT)\" 
-L/usr/cisco/packages/libxslt/libxslt-1.1.6/lib 
-L/usr/cisco/packages/libxml2/libxml2-2.6.9/lib     -lxslt -lxml2 -lz -lpthread 
-lm -L/usr/cisco/packages/libxml2/current/lib -lxml2 -lz -lpthread -lm  
conftest.c -lssh2   >&5
360 /usr/bin/ld: skipping incompatible 
/usr/cisco/packages/libxslt/libxslt-1.1.6/lib/libxslt.a when searching for 
-lxslt
361 /usr/bin/ld: cannot find -lxslt




I tried with --disable-libssh2 option and i see : 
checking for libssh2_version in -lssh2... no
configure: error: Missing libssh2.

Contents of config.log : 
configure:12169: gcc -o conftest 
-I/usr/cisco/packages/libxslt/libxslt-1.1.6/include 
-I/usr/cisco/packages/libxml2/current/include/libxml2 -I/usr/cisco/packa     
ges/libxml2/current/include/libxml2  -O3   -DRCSID=\"$(IDGIT)\" 
-DDISABLE_LIBSSH -L/usr/cisco/packages/libxslt/libxslt-1.1.6/lib 
-L/usr/cisco/packages/libxml     2/libxml2-2.6.9/lib -lxslt -lxml2 -lz 
-lpthread -lm -L/usr/cisco/packages/libxml2/current/lib -lxml2 -lz -lpthread 
-lm  conftest.c -lpthreads  -lutil >&5
 363 /usr/bin/ld: skipping incompatible /usr/cisco/packages/libxslt/libxslt-1.1.6/lib/libxslt.a when searching for -lxslt
 364 /usr/bin/ld: cannot find -lxslt
 365 collect2: ld returned 1 exit status
 366 configure:12169: $? = 1
 367 configure: failed program was:

On my system i see that there is path to libxslt

 ll /usr/lib/libxslt.so.1*
lrwxrwxrwx  1 root root     17 Oct 18  2011 /usr/lib/libxslt.so.1 -> 
libxslt.so.1.1.11
-rwxr-xr-x  1 root root 192608 Jul 23  2008 /usr/lib/libxslt.so.1.1.11



What version of the product are you using? On what operating system?
 0.6.0 on LINUX

Please provide any additional information below.

What steps will reproduce the problem?
1. Connect to libnetconf based server.
2. send create-subscription request without any parameters.


What is the expected output?
As stated in section 2.1.1 of RFC 5277 
(http://tools.ietf.org/html/rfc5277#section-2.1.1): If <startTime> is not 
present, this is not a replay subscription.

What do you see instead?
Whole stream is replayed.

Please use labels and text to provide additional information.

What steps will reproduce the problem?
1. run edit-config operation without test-option

What is the expected output? What do you see instead?

According to RFC 6241 sec. 7.2, test-then-set is default test-option if the 
server supports :validate capability. But libnetconf does no validation if 
test-option is not specified.

What steps will reproduce the problem?
1. Connect to server.
2. Send message with equal source and target datastore, for example:
<rpc xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="35">
  <copy-config>
        <source>
                <candidate/>
        </source>
        <target>
                <candidate/>
        </target>
</copy-config>
</rpc>


What is the expected output?
If the <source> and <target> parameters identify the same URL or
configuration datastore, an error MUST be returned with an error-
tag containing "invalid-value". [RFC6241]

<rpc-reply xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="35">
  <rpc-error>
    <error-tag>invalid-value</error-tag>
    <error-type>protocol</error-type>
    <error-severity>error</error-severity>
    <error-message>The request specifies an unacceptable value for one or more parameters.</error-message>
  </rpc-error>
</rpc-reply>

What do you see instead?

<rpc-reply xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="35">
  <ok/>
</rpc-reply>


Please use labels and text to provide additional information.

What steps will reproduce the problem?
1. Use two or more individual yin schema;

What is the expected output? What do you see instead?
At initializing stage, the yinmodel_parse fails to update the prefix tag.

What version of the product are you using? On what operating system?
Master branch, one month old. Ubuntu.

Please provide any additional information below.
Fixed(in my code) in yinmodel_parse routine using:
for (i=0; ns_mapping[i] != NULL; i=i+3) {
    if (strcmp(ns_mapping[i+1], yin->ns_uri) == 0) {
        yin->ns_prefix = strdup(ns_mapping[i+2]);
        break;
    }
}

What steps will reproduce the problem?
1. Start the server and client
2. Send a edit-config operation with error-option set to "continue-on-error" 
and three configurations (that will call three callbacks) where the second 
callback return EXIT_FAILURE.

What is the expected output? What do you see instead?
I expected the three callbacks should be called but the behaviour is identical 
to "stop-on-error" error-option (just the first and second callbacks are 
called).

What version of the product are you using? On what operating system?
I am using the branch master commit 3f33930cd3a8 (Thu Sep 5 13:50:40 2013 
+0200) on a Gentoo Linux.

What steps will reproduce the problem?
1. change datastore file
2. start NETCONF server
3. connect with client - running does not match startup from file

What is the expected output? What do you see instead?
I tried to disable nacm, however, it is still enabled

Please use labels and text to provide additional information.
built from git fde95fffdb9af2a04087

What steps will reproduce the problem?

Just add more (leaf-)list nodes into a configuration without any.

What is the expected output? What do you see instead?

Only the first (leaf-)list is correctly handled, the rest is completely ignored.

Problem was found by performing "bad" edit-config XML.
What steps will reproduce the problem?
1. Content of datastore:
<nacm xmlns="urn:ietf:params:xml:ns:yang:ietf-netconf-acm">
 <rule-list>
  <rule>
   <name>permit-all-last</name>
   <module-name>*</module-name>
   <access-operations>*</access-operations>
   <action>permit</action>
  </rule>
  <rule>
   <name>permit-all-last2</name>
   <module-name>*</module-name>
   <access-operations>*</access-operations>
   <action>permit</action>
  </rule>
 </rule-list>
</nacm>

2. edit-config with the wrong key attribute that tries to move item "before 
itself":

<nacm xmlns="urn:ietf:params:xml:ns:yang:ietf-netconf-acm" 
xmlns:yang="urn:ietf:params:xml:ns:yang:1">
<rule-list>
<rule yang:insert="before" yang:key="[name='permit-all-last']">
<name>permit-all-last</name>
</rule>
</rule-list>
</nacm>

What is the expected output? What do you see instead?
In the output permit-all-last is completely missing.

Implement handling of kill-session rpc according RFC 6241.

As Paulo Zulato suggested in mailing list; It will be useful to have chance to 
report error details from transapi module to client.

What steps will reproduce the problem?
1) netconf-client - connect to server
2) user-rpc with this:
<edit-config>
<target>
<running/>
</target>
<config>
<comet-testers xmlns="http://cesnet.cz/ns/yang/comet-tester">
<comet-tester>
<interface-id>0</interface-id>
<statistics>
<enabled>false</enabled>
</statistics>
</comet-tester>
</comet-testers>
</config>
</edit-config>

OR
edit-config with:
<comet-testers xmlns="http://cesnet.cz/ns/yang/comet-tester">
<comet-tester>
<interface-id>0</interface-id>
<statistics>
<enabled>false</enabled>
</statistics>
</comet-tester>
</comet-testers>


What is the expected output? What do you see instead?
NETCONF error: operation-failed (application) - The request requires a resource 
that already is in use.

Please use labels and text to provide additional information.
get, get-config works correctly

Hello, i'm an university student and i would like to use libnetconf for my 
college internship; our goal is to show how Netconf can be used to manipulate a 
network device; do libnetconf provide all the instruments required for this? Is 
the Server partially implemented? What kind of device can act as Netconf 
Server? A Linux or Windows PC can act as Netconf Server? Thank you so much for 
your help!

What steps will reproduce the problem?
1. Start the client with a empty running configuration and the attached module
2. Send the messages of the "test1" or "test2" attachments.

What is the expected output? What do you see instead?
The expected outputs and the obtained outputs for messages of the "test1" and 
"test2" attachments are in the "result1" and "result2" attachments, 
respectively. Briefly, when I replace three list nodes only the first callback 
is called (at the branch master). When there is a leaf before the lists only 
this callback is called.

What version of the product are you using? On what operating system?
I am using the branch master commit abc9e2df38cb (Mon Sep 9 11:26:25 2013 
+0200) and branch transapi_change commit 0c2e4c73d9e (Tue Sep 10 14:59:18 2013 
+0200) on a Gentoo Linux.

If YANG ordered-by-user element is used and the content of a configuration file 
is not changed, only the order of the items, then transapi will fail to detect 
the change. The desired behaviour is for every moved item to generate first 
removal and then addition.

What steps will reproduce the problem?
1.
Yang File:
....
 feature damagedBooks {
       description
        "Testing if-feature using damagedBooks feature.";
       reference
        "RFC 6020: testing if-feature";
     }

...

leaf damaged-books {
            if-feature damagedBooks; <<<<
            type uint32;
      }


2.
applying the configuration data on the client like below

<toaster xmlns="http://netconfcentral.org/ns/toaster"  
xmlns:xc="urn:ietf:params:xml:ns:netconf:base:1.0" xc:operation="create">
<damaged-books>11</damaged-books>
</toaster>

is getting accepted despite the fact that for the leaf we have a check of 
"if-feature" for damagedBooks.  By default all features are disabled by 
libnetconf.

3.
If we used the configuration data to give a new value using merge like below. 

<toaster xmlns="http://netconfcentral.org/ns/toaster"  
xmlns:xc="urn:ietf:params:xml:ns:netconf:base:1.0" xc:operation="merge">
<damaged-books>10</damaged-books>
</toaster>

This throws an error as below. 
NETCONF error: unknown-element (application) - An unexpected element is present.

The same error should have thrown for configuration applied in the step2. 

What is the expected output? What do you see instead?
The error in step 3 should have appeared for step 2 also. The feature is 
getting check for "merge" not while creating. 

What version of the product are you using? On what operating system?

Version :
---------
libnetconf 0.8.0

Operating System :
-------------------
Xbuntu 14.04 LTS

Please provide any additional information below.

What steps will reproduce the problem?
1.just start netopeer-cli and try to connect, process stuck in password getting

What is the expected output? What do you see instead?
Scan password correctly


What version of the product are you using? On what operating system?
22783e3377da

Please provide any additional information below.

The function callback_sshauth_password_default() is not working properly for 
big endian systems. The function is running in the read loop forever.
Attached you will find the patch file.

I'm trying to establish a Netconf session with a Juniper MX480 router. I'm 
getting errors during initial negotiation.

I have followed these steps:

1. Open client example with:

$ netconf-client 


2. Establish verbosity level with:

netconf> verbose
Verbose level set to VERBOSE


3. Connect and authenticate to the Juniper MX480 router:

netconf> connect --login <user> <ip_address> 
libnetconf VERBOSE: Keyboard-interactive authentication
libnetconf ERROR: Authentication failed (Authentication failed 
(keyboard-interactive))
libnetconf VERBOSE: Password authentication (host <ip_address>, user <user>)
<user>@<ip_address> password: 


4. Getting the following errors:

libnetconf ERROR: No base capability found in the capabilities intersection.
libnetconf ERROR: Unsupported NETCONF protocol version (-1)
libnetconf ERROR: Malformed message received, closing the session 22273.
connect: connecting to the <ip_address> failed.
netconf> 


I don't know how to get logging of my session to see actual Netconf messages, 
but I was using netopeer and these was my issue:
http://code.google.com/p/netopeer/issues/detail?id=8&can=1

steps to reproduce:
1. send the following rpc:
<?xml version="1.0" encoding="UTF-8"?>
<nc:rpc xmlns:nc="urn:ietf:params:xml:ns:netconf:base:1.0"
    message-id="1">
    <nc:edit-config>
        <nc:target>
            <nc:running />
        </nc:target>
        <nc:config>
            <config />
        </nc:config>
    </nc:edit-config>
</nc:rpc>

(I used ncclient [1] to send the request); backtrace is below.

The attached patch adds a check to prevent the segfault in 
ncdflt_edit_remove_default.


best
Tobi


[1] http://ncclient.grnet.gr/

the backtrace:
(gdb) bt
#0  0x00007ffff7b899eb in ncdflt_edit_remove_default (config=0x87be70, 
model=0x7697f0) at src/with_defaults.c:772
#1  0x00007ffff7b9bc5a in ncds_apply_rpc (id=1681692778, session=0xb89ef0, 
rpc=0x971fb0) at src/datastore.c:5376
#2  0x00007ffff7b9d5da in ncds_apply_rpc2all (session=0xb89ef0, rpc=0x971fb0, 
ids=0x0) at src/datastore.c:5955
#3  0x0000000000403dad in server_process_rpc (session=0xb89ef0, rpc=0x971fb0) 
at src/server_operations.c:563
#4  0x0000000000406f08 in process_operation (socket=17) at 
src/server_comm_socket.c:396
#5  0x000000000040729a in comm_loop (conn=0x60aa08 <sock>, timeout=500) at 
src/server_comm_socket.c:494
#6  0x000000000040391e in main (argc=3, argv=0x7fffffffe3d8) at src/server.c:273

What steps will reproduce the problem?
After libnetconf update, webgui deadlocks after NETCONF request.
Before this deadlock, many replayComplete messages appeared:
Received message (session 14010): <?xml version="1.0"?>\n<notification 
xmlns="urn:ietf:params:xml:ns:netconf:notification:1.0">\n  
<eventTime>2013-11-26T15:22:35Z</eventTime>\n  <replayComplete 
xmlns="urn:ietf:params:xml:ns:netmod:notification"/>\n</notification>\n


Libnetconf was updated by David Kupka using /libnetconf-0.6.99-2.x86_64.

Feature description:

Prepare client-side changes to support transport over TLS 
([http://tools.ietf.org/html/draft-badra-netconf-rfc5539bis-02 RFC 5539bis]) 
and describe necessary settings for the NETCONF server based on libnetconf.

Feature description:

Create YANG data model implemented by the library with its own enhancements.

As first, libnetconf now generates notifications on datastore lock/unlock. The 
notification is under "urn:cesnet:params:xml:ns:libnetconf:notifications " 
namespace. The data model will be announced by the libnetconf-based servers in 
<hello> messages.

Feature description:

Support default values defined inside the typedef statements.