DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://nagoya.apache.org/bugzilla/show_bug.cgi?id=6362>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=6362 Multithreading problems during regeneration of sitemap Summary: Multithreading problems during regeneration of sitemap Product: Cocoon 2 Version: Current CVS Platform: All OS/Version: All Status: NEW Severity: Critical Priority: Other Component: core AssignedTo: [EMAIL PROTECTED] ReportedBy: [EMAIL PROTECTED] Hi, I just found out this problem with the asynchronous sitemap generation: If you change the sitemap in a running system the sitemap is regenerated in the background. In the meantime all incomming requests are served with the old sitemap until the new one is finished. This is nice. But, if a new request enters the old sitemap and is in the processing of this sitemap and if at the same time the regeneration is finished, the old sitemap is released by the sitemap.handler class. >> taken from the run() method << if (this.sitemap != null) { programGenerator.release((CompiledComponent)this.sitemap); } << This release causes a dispose on the Sitemap object which in turn resets the sitemap object and for example sets the manager to null. But the new requests is currently processing with the old sitemap object which then results sooner or later with a NPE (as the manager is now null). This is easily reproducable. just change the sitemap and hit several times on reload. Now, I see two solutions: a) Each request looksup and releases the sitemap, so the release in the run() method of the handler will not dispose the old sitemap until all requests still using that old sitemap are finished. b) Remove the asynchron regeneration. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
