[NETCONF-208] Controller-config still fails to operate Created: 23/May/16  Updated: 15/Mar/19  Resolved: 25/May/16

Status: Resolved
Project: netconf
Component/s: netconf
Affects Version/s: None
Fix Version/s: None

Type: Bug
Reporter: Vratko Polak Assignee: Jakub Morvay
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: All
Platform: All

Issue Links:
Blocks
is blocked by YANGTOOLS-614 master fails mounting controller-config Resolved
Duplicate
duplicates YANGTOOLS-614 master fails mounting controller-config Resolved
External issue ID: 5943

 Description   

Even after fix [0] to Boron YANGTOOLS-614 was merged, system tests are still failing.
Both on initial GET [1] and on subsequent PUT [2].

[0] https://git.opendaylight.org/gerrit/39167
[1] https://jenkins.opendaylight.org/releng/view/netconf/job/netconf-csit-1node-userfeatures-only-boron/337/robot/report/log.html#s1-s1-s1-t1-k3-k1-k57-k2
[2] https://jenkins.opendaylight.org/releng/view/netconf/job/netconf-csit-1node-userfeatures-only-boron/337/robot/report/log.html#s1-s3-s1-t3-k2-k2-k1-k3-k7-k1

 Comments   
Comment by Tomas Cere [ 23/May/16 ]

This is still a symptom of the original issue in 5917 that karaf upgrade caused.
Apart from the javax.xml.transform.TransformerFactory thats mentioned in 5917 javax.xml.parsers.DocumentBuilderFactory was also added into karaf properties which is causing namespace inheritance issues in cfg subsystem.

Comment by Vratko Polak [ 23/May/16 ]

In 3-node tests, device is created using odl-netconf-clustered-topology, and that works without issues.

Comment by Vratko Polak [ 23/May/16 ]

> In 3-node tests, device is created

Correction: Only testtool devices work. Attempt to read from config subsystem northbound fails the same way as in 1-node tests:
https://jenkins.opendaylight.org/releng/view/bgpcep/job/bgpcep-csit-3node-periodic-bgpclustering-only-boron/6/robot/report/log.html#s1-s2-t5-k3-k1-k1-k1-k2-k10-k1-k2-k4-k1

Comment by Vratko Polak [ 24/May/16 ]

Here is what the exception looks like, even with https://git.opendaylight.org/gerrit/#/c/39368/1 applied:

2016-05-24 17:54:14,894 | WARN | oupCloseable-3-1 | KeepaliveSalFacade | 281 - org.opendaylight.netconf.sal-netconf-connector - 1.4.0.SNAPSHOT | RemoteDevice

{controller-config}

: Rpc failure detected. Reconnecting netconf session
java.lang.IllegalStateException: Unknown child(ren) node(s) detected, identified by: module, in: container modules
at com.google.common.base.Preconditions.checkState(Preconditions.java:197)[65:com.google.guava:18.0.0]
at org.opendaylight.yangtools.yang.data.impl.schema.SchemaUtils.findSchemaForChild(SchemaUtils.java:115)[85:org.opendaylight.yangtools.yang-data-impl:1.0.0.SNAPSHOT]

What is the correct value for javax.xml.parsers.DocumentBuilderFactory property?

Comment by Vratko Polak [ 25/May/16 ]

Not even https://git.opendaylight.org/gerrit/#/c/39368/2 fixes this.
I am not sure which debug logs to enable to see why.

Comment by Vratko Polak [ 25/May/16 ]

The following changes seem so fix this Bug:
https://git.opendaylight.org/gerrit/#/c/39368/5 (for distributions based on karaf-parent)
https://git.opendaylight.org/gerrit/#/c/39402/1 (for integration/distribution)

Comment by Jakub Morvay [ 25/May/16 ]

Closing this bug as this is duplicate of YANGTOOLS-614 since this is related to issues with downgrade of behaviour of javax.xml.transform.TransformerFactory implementation.

org.apache.xalan.processor.TransformerIdentityImpl from org.apache.servicemix.bundles.xalan-2.7.2_2.jar, seems to have issues with nemaspace defaulting when transforming SAX events to DOM source.

As Vratko pointed out, patches
https://git.opendaylight.org/gerrit/#/c/39368/5 (for distributions based on karaf-parent)
https://git.opendaylight.org/gerrit/#/c/39402/1 (for integration/distribution)

should fix this bug.

Generated at Wed Feb 07 20:14:26 UTC 2024 using Jira 8.20.10#820010-sha1:ace47f9899e9ee25d7157d59aa17ab06aee30d3d.