Runtime only resources
by Michael Musgrove
We have a transaction case where the customer is trying to delete a
(transaction log) model resource via the CLI and is seeing the error:
"was directly invoked by the user. User operations are not permitted to
directly update the persistent configuration of a server in a managed
domain."
This is coming from ControllerLogger#modelUpdateNotAuthorized (in
wildfly-core) because we define the resource as *not* being runtime only:
public boolean isRuntime() {return false;}
My question is should we be defining the resource as a runtime resource or
should we tell the customer not to delete the resouce from the controller?
Thanks,
Mike
--
Michael Musgrove
Transactions Team
e: mmusgrov(a)redhat.com
Our mission:To be the catalyst in communities of customers, contributors,
and partners creating better technology the open source way.
JBoss, by Red Hat
Registered Address: Red Hat Ltd, 6700 Cork Airport Business Park, Kinsale
Road, Co. Cork.
Registered in the Companies Registration Office, Parnell House, 14 Parnell
Square, Dublin 1, Ireland, No.304873
Directors:Michael Cunningham (USA), Vicky Wiseman (USA), Michael O'Neill,
Keith Phelan, Matt Parson (USA)
8 years, 5 months
Upgrade to WildFly Elytron 1.1.0.Beta54
by Darran Lofthouse
For any components with a dependency on WildFly Elytron I would recommend
upgrading to the 1.1.0.Beta54 release which was released last week to
double check all API usage is up to date.
We don't expect many more changes now as we are heading into the candidate
release this week but if any changes are required we will e-mail this list
first.
One change we do know is on it's way is next week we will be removing
PrincipalDecoder from the project entirely, this is one of the earliest
APIs we added but has now been superseded by the principal transformers
represented by Function<Principal, Principal>.
Regards,
Darran Lofthouse.
8 years, 5 months
AuditLog Test failures/JBoss Modules.
by Kabir Khan
See below. I have been seeing this a bit recently, At least the JMXAuditLogTestCase failure is due to that JBoss Modules has a WeakReference to the ServiceLoaders. When the weak reference is cleared this results (via the WeakReference's ReferenceQueue and the ModuleLoader's Reaper) in a call to unregister the MBean associated with the ServiceLoader, which shows up as an additional unexpected entry in the AuditLog. This clearing of the weak reference is unpredictable and can happen at any time.
I plan to fix this test test today, and to add a better message to the AuditLogBootingSyslogTest.
Example failures:
https://ci.wildfly.org/viewLog.html?buildId=65226&buildTypeId=WildFlyCore...
https://ci.wildfly.org/viewLog.html?buildId=65225&buildTypeId=WildFlyCore...
> On 22 Jun 2017, at 00:55, James R. Perkins <notifications(a)github.com> wrote:
>
> Hmm... ...kind of weird that two log related tests failed however these are audit log tests which this should touch at all. retest this please.
>
> —
> You are receiving this because you are subscribed to this thread.
> Reply to this email directly, view it on GitHub, or mute the thread.
>
8 years, 6 months
WildFly Client Configuration and schema
by David M. Lloyd
Something that has come up in a JIRA issue [1] is the question of
whether the WildFly Client Configuration file should use a schema for
its root element.
The root of the "wildfly-client.xml" file is presently a single,
namespace-less element named "configuration". The content of this
element is a sequence of library-specific configuration, like this:
<configuration>
<discovery xmlns="urn:wildfly-discovery:1.0">
<!-- content here -->
</discovery>
<jboss-ejb-client xmlns="urn:jboss:wildfly-client-ejb:3.0">
<!-- content here -->
</jboss-ejb-client>
<!-- etc. -->
<configuration>
The question is, should there be a schema for that root element? The
entire schema would be something like:
<xs:element name="configuration" type="configuration-type"/>
<xs:complexType name="configuration-type">
<xs:any maxOccurs="unbounded" namespace="##other"/>
</xs:complexType>
Then you'd have to put the configuration namespace in as well:
<configuration xmlns="urn:wildfly-client:1.0">
<discovery xmlns="urn:wildfly-discovery:1.0">
<!-- content here -->
</discovery>
<jboss-ejb-client xmlns="urn:jboss:wildfly-client-ejb:3.0">
<!-- content here -->
</jboss-ejb-client>
<!-- etc. -->
</configuration>
WDYT? Is it worth having a whole schema/namespace for a single element?
[1] https://issues.jboss.org/browse/ELY-1098
--
- DML
8 years, 6 months
WildFly's HipChat room
by Tristan Tarrant
Hi all,
I've complained in the past about the WildFly's team decision to move to
HipChat, but I've always been able to access the chat room as a guest
using the link provided on the homepage [1]
Unfortunately that link no longer works, since it expects me to register
with an e-mail account. Considering HipChat's stupid policy that
disallows a single account to join multiple groups, I'm unwilling to
"sacrifice" an e-mail address just for the few situations in which I
need to get hold of one of you.
Please fix it.
Tristan
[1] https://www.hipchat.com/gSW9XYz69
--
Tristan Tarrant
Infinispan Lead
JBoss, a division of Red Hat
8 years, 6 months
