[
https://issues.apache.org/jira/browse/SLING-4352?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dirk Rudolph updated SLING-4352:
--------------------------------
Comment: was deleted
(was: This is an urgent issue at least with AEM 6.1 base on Servlet Resolver
2.3.6.
It happens that when the internal Cache of the Servlet resolver is flushed and
multiple concurrent requests are handled by the same Servlet/Script the
perThreadResolver is reused in different Threads and causes an ISE within
ResourceResolver#checkClosed when the request that initially got the
ScriptResource has been finished but the perThreadResolver is still in the
ScriptContext of another Thread.
Maybe thats not the same issue but kind of related?)
> The ResourceResolver passed to SlingScripts should not be closeable
> -------------------------------------------------------------------
>
> Key: SLING-4352
> URL: https://issues.apache.org/jira/browse/SLING-4352
> Project: Sling
> Issue Type: Bug
> Components: Servlets
> Reporter: Radu Cotescu
> Fix For: Servlets Resolver 2.3.8
>
>
> The {{SlingServletResolver}} uses an administrative resource resolver
> ({{sharedScriptResolver}}) for script resolution. This same resolver is
> passed to SlingScripts when a resource is adapted to a {{SlingScript}} in
> {{SlingServletResolver#findScript}} (see
> {{SlingScriptAdapterFactory#getAdapter}}).
> Since this resolver is made available to scripts in the {{ScriptContext}}
> (see the implementation of {{DefaultSlingScript#call}}), the resolver should
> be protected against accidental closing from downstream callers.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
