[JBoss JIRA] (WFLY-12900) Clustering TS: Default to non-multicast discovery protocol
by Radoslav Husar (Jira)
[ https://issues.redhat.com/browse/WFLY-12900?page=com.atlassian.jira.plugi... ]
Radoslav Husar updated WFLY-12900:
----------------------------------
Affects Version/s: 19.0.0.Beta1
> Clustering TS: Default to non-multicast discovery protocol
> ----------------------------------------------------------
>
> Key: WFLY-12900
> URL: https://issues.redhat.com/browse/WFLY-12900
> Project: WildFly
> Issue Type: Task
> Components: Clustering, Test Suite
> Affects Versions: 18.0.1.Final, 19.0.0.Beta1
> Reporter: Radoslav Husar
> Assignee: Radoslav Husar
> Priority: Critical
>
> We have been slowly making progress on reducing reliance on multicast to run the clustering test suite by default given the complexity of system configuration (e.g. multicast on {{::1}} workarounds). The goal is to use a default non-multicast-based protocol by defaut.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
6 years, 3 months
[JBoss JIRA] (DROOLS-2254) Automate .proto files rebuild in pom.xml
by Michael Biarnes Kiefer (Jira)
[ https://issues.redhat.com/browse/DROOLS-2254?page=com.atlassian.jira.plug... ]
Michael Biarnes Kiefer updated DROOLS-2254:
-------------------------------------------
Fix Version/s: 7.33.0.Final
(was: 7.32.0.Final)
> Automate .proto files rebuild in pom.xml
> ----------------------------------------
>
> Key: DROOLS-2254
> URL: https://issues.redhat.com/browse/DROOLS-2254
> Project: Drools
> Issue Type: Task
> Components: tools
> Affects Versions: 7.5.0.Final
> Reporter: Dmitry Volodin
> Assignee: Dmitry Volodin
> Priority: Minor
> Fix For: 7.33.0.Final
>
>
> According to contribution guide, any .proto file or protobuf version changes it's necessary to download protoc utility and regenerate Java classes based on .proto files.
> This will add automation for downloading protoc utility and Java classes generation based on Maven Protocol Buffers Plugin. There is no timestamps and other build related info inside generated Java files and no changes will be added on each new build.
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
6 years, 3 months
[JBoss JIRA] (WFLY-13009) moduleAvailability message is sent before module has started
by Richard Achmatowicz (Jira)
[ https://issues.redhat.com/browse/WFLY-13009?page=com.atlassian.jira.plugi... ]
Richard Achmatowicz updated WFLY-13009:
---------------------------------------
Description:
When EJBs are deployed in a clustered server environment, moduleAvailability messages are sent from the server to all connected EJB client applications in order to inform the client that the module is available on that node.
The sending of moduleAvailability messages involves the DeploymentRepository, which records which modules are deployed, and the ModuleAvailabilityListener, which is a DeploymentRepositoryListener which gets called at various stages of module deployment (adding, starting, stopping).
It looks as though moduleAvailability message are being sent earlier than they should be; in other words, sent when the module is added to the repository and not when the module is later started. This can cause client invocations on the module to return a NoSuchEJBException even after the moduleAvailability message has been sent to the client, causing what is seen in EJBCLIENT-362.
was:
When EJBs are deployed in a clustered server environment, moduleAvailability messages are sent from the server to all connected EJB client applications in order to inform the client that the module is available on that node.
The sending of moduleAvailability messages involves the DeploymentRepository, which records which modules are deployed, and the ModuleAvailabilityListener, which is a DeploymentRepositoryListener which gets called at various stages of module deployment (adding, starting, stopping).
It looks as though moduleAvailability message are being sent earlier than they should be; in other words, sent when the module is added to the repository and not when the module is later started. This can cause client invocations on the module to return a NoSuchEJBException even after the moduleAvailability message has been sent to the client, causing what is seen in EJBCLIENT-362.
> moduleAvailability message is sent before module has started
> ------------------------------------------------------------
>
> Key: WFLY-13009
> URL: https://issues.redhat.com/browse/WFLY-13009
> Project: WildFly
> Issue Type: Bug
> Components: Clustering, EJB
> Affects Versions: 19.0.0.Beta1
> Reporter: Richard Achmatowicz
> Assignee: Paul Ferraro
> Priority: Major
>
> When EJBs are deployed in a clustered server environment, moduleAvailability messages are sent from the server to all connected EJB client applications in order to inform the client that the module is available on that node.
> The sending of moduleAvailability messages involves the DeploymentRepository, which records which modules are deployed, and the ModuleAvailabilityListener, which is a DeploymentRepositoryListener which gets called at various stages of module deployment (adding, starting, stopping).
> It looks as though moduleAvailability message are being sent earlier than they should be; in other words, sent when the module is added to the repository and not when the module is later started. This can cause client invocations on the module to return a NoSuchEJBException even after the moduleAvailability message has been sent to the client, causing what is seen in EJBCLIENT-362.
>
--
This message was sent by Atlassian Jira
(v7.13.8#713008)
6 years, 3 months