<html><head><meta http-equiv="Content-Type" content="text/html charset=windows-1252"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Folks,<div><br></div><div>we are still having disconnects between the OSGi Bundle lifecycle operations and their respective mapping to the deployment API. Here I'd like to describe how it currently works and how it could possibly get fixed. Related issues are tracked by <a href="https://issues.jboss.org/browse/AS7-5567">[1] </a>- further background is described in <a href="https://docspace.corp.redhat.com/docs/DOC-139013">[2]</a>.<br><div><br></div><div>Bundle start/stop disconnect</div><div>--------------------------------------</div><div><br></div><div>A Bundle start/stop operation activates/deactivates the bundle services. The associated class loader and wiring is unaffected.</div><div><br></div><div>Approach #1 - start/stop is not mapped to deployment operations</div><div><br></div><div><ul class="MailOutline"><li>Bundle deploy installs the Bundle to the Framework</li><li>Bundle is not resolved/started as part of the deploy operation</li><li>No module/classloader is created during deployment</li><li>All deployment processing that needs load classes/resources is skipped</li><li>No service provided by the bundle deployment is created</li><li>Bundle needs to be started/stopped explicitly </li><li>Bundle start/stop management operations are specific to bundle deployments</li><li>The POST_MODULE processing for OSGi enabled deployments (e.g. webapp) would need to be executed in an OSGi specific processor chain</li><li>OSGi enabled POST_MODULE DUPs have their functionality externalised such that it can get reused for start/stop operations</li><li>The start/stop processing chain must be traversable in both directions repeatedly</li></ul><div><br></div></div><div>Approach #2 - start/stop is mapped to deployment operations</div><div><br></div><div><ul class="MailOutline"><li>Bundle deploy installs the Bundle to the Framework and attempts to resolve/start the Bundle</li><li>If the Bundle can get resolved, a module/classloader is created as part of the DU processing</li><li>In case the Bundle cannot get resolved, DUP processing is deferred and MODULE phase processing is reattempted later based on an external trigger</li><li>Has the benefit that any deployment type can also be a Bundle</li><li>Bundle stop must reverse work that is done in POST_MODULE DUPs</li><li>POST_MODULE DUPs must be traversable in both direction multiple times</li></ul><div><br></div></div><div>Currently, we use approach #2 with limited support from POST_MODULE DUPs (i.e. they are not designed to be executed multiple times for the same deployment). There is also the cleanup phase which breaks multiple executions of POST_MODULE DUPs.</div><div><br></div><div><div>Bundle uninstall disconnect</div><div>--------------------------------------</div></div><div><br></div><div>Bundle uninstall only removes the BundleRevision from the runtime if it is no longer in use. The classloader associated with an uninstalled bundle that is still in use remains active.</div><div><br></div><div>Approach #1 - "in use" semantic specific to Bundle deployments</div><div><br></div><div><ul class="MailOutline"><li>An undeploy management operation undeploys the bundle revision but holds in the MODULE phase when the revision is still in use</li><li>This is similar to deferred MODULE phase for deploy operation</li><li>in use is not respected for non-osgi deployments and hence breaks bundle deployments that depend on them</li></ul></div><div><br></div><div><div>Approach #2 - "in use" semantic applies to all deployment types</div></div><div><br></div><div><ul class="MailOutline"><li>Undeploy will only destroy the Module service if the Module is not in use by another deployment</li><li>When Module usage drops to zero, undeploy continues and removes the deployment from management </li><li>This approach would map cleanly - there is no cascading destroy effect. i.e. when you undeploy a jar that a webapp depends on</li><li>Services provided by the undeployed deployment would always go down, only classloader destruction is deferred</li></ul></div><div><div><br class="webkit-block-placeholder"></div><div><div><div>Bundle update disconnect</div><div>------------------------------------</div></div></div><div><br></div><div>A Bundle can have multiple revisions. A call to Bundle.update(…) creates a new revision. There is only ever one current revision (the latest update) associated with a Bundle. Stale revisions continue to exist in the runtime until an explicit refresh of the transitive dependency graph. </div><div><br></div><div><div>Approach #1 - bundle revisions are represented as individual deployments</div></div><div><br></div><div><ul class="MailOutline"><li>Each bundle update deploys the revision as a new deployment</li><li>The association of revisions to bundle happens in the OSGi layer only</li><li>Management sees disconnected deployments with no notion of stale/current or belonging together</li></ul></div><div><br></div><div><div>Approach #2 - deployment API supports the notion of multiple revisions</div></div><div><br></div><div><ul class="MailOutline"><li>deployments, their revisions and respective mapping would be modelled more closely on OSGi requirements</li><li>some deployment types might not support revisions </li></ul><div><br></div></div><div>------------------------------------------------------------------------------------------------------------------------------------------------------</div><div><br></div><div>I suspect that solving these disconnects can happen in the context of AS9 only.</div><div><br></div><div>Looking forward to your feedback</div><div><br></div><div>Cheers</div><div>--thomas</div><div><br></div><div><div><a href="https://issues.jboss.org/browse/AS7-5567">[1]</a> [AS7-5567] Resolve AS8/OSGi disconnects </div></div><div><a href="https://docspace.corp.redhat.com/docs/DOC-139013">[2]</a> AS/OSGi Deployment Lifecycle Requirements</div><div>
<div><pre class="moz-signature" cols="72">xxxxxxxxxxxxxxxxxxxxxxxxxxxx
Thomas Diesler
JBoss OSGi Lead
JBoss, a division of Red Hat
xxxxxxxxxxxxxxxxxxxxxxxxxxxx </pre><div><br></div></div><br class="Apple-interchange-newline">
</div>
<br></div></div></body></html>