Return-Path: Delivered-To: apmail-xml-cocoon-dev-archive@xml.apache.org Received: (qmail 2949 invoked by uid 500); 29 Jul 2001 11:16:34 -0000 Mailing-List: contact cocoon-dev-help@xml.apache.org; run by ezmlm Precedence: bulk list-help: list-unsubscribe: list-post: Reply-To: cocoon-dev@xml.apache.org Delivered-To: mailing list cocoon-dev@xml.apache.org Received: (qmail 2938 invoked from network); 29 Jul 2001 11:16:34 -0000 From: "Carsten Ziegeler" To: Subject: AW: [C2B2] Big Bug in sub-sitemap handling (?) Date: Sun, 29 Jul 2001 13:17:05 +0200 Message-ID: MIME-Version: 1.0 X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2911.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 In-Reply-To: <001701c1178a$1b61bb40$01c8c8c3@RAEL> X-MIMETrack: Itemize by SMTP Server on PBSN1/Systeme und Netzwerke(Release 5.0.5 |September 22, 2000) at 29.07.2001 13:16:31, Serialize by Router on PBSN1/Systeme und Netzwerke(Release 5.0.5 |September 22, 2000) at 29.07.2001 13:16:32 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable X-Spam-Rating: h31.sny.collab.net 1.6.2 0/1000/N Ok, I was wrong... The bug was in the SitemapComponentSelector which didn't call the parent selector when a component was released. I fixed this, but I am not sure if this is a correct solution: The component is released in the selector itself AND in the parent selector as we can't detect which selector was used to lookup the component. Is this ok, Berin? But I still think that my first guess is correct, too. The invoke() method with the pipeline objects is of course only called for internal pipelines. So my guess is still that the releasing scheme doesn't work for internal pipelines in subsitemaps (when called from the top sitemap). Any experience here? Now I can go to the beach again... Carsten > Carsten Ziegeler wrote: > > Hi, > > as I am on vacation right now, I can't test or even build c2. Looking > through the generated code > for subsitemaps for about 5 minutes, I think it is wrong: > > When a subsitemap is entered the invoke() method of the sitemap manager is > called, this > method gets among some other paramters the EventPipeline and the > StreamPipeline. These are > passed to the subsitemap and used there to initialize components which > belong to a > ComponentManager of the subsitemap. After the event is processed, the root > sitemap > releases the objects with the ComponentManager of the root sitemap and not > of the subsitemap, > so they cannot be released. > This actually is only my guess. Perhaps I am wrong. (But I hope not). > > So, could someone test this by simply changing the sitemap.xsl, > so that the > invoke() method > of the sitemap manager is called without passing the event and stream > pipeline, please? The > manager has the invoke method with both signatures. > > If I get some free time in my vacation I could perhaps look at the result > tomorrow evening. > But don't expect any reply from me for the next days...Sorry! > > Greetings from the sun, > > Carsten > > >>>>>>>>>>> > Matthew Langham wrote: > > After comparing the logs I can now add a little to that: > > When the pipeline is in the subsitemap components such as the > TraxTransformer are =5Fretrieved=5F from the pool - but they are not retu= rned! > This is the case if the pipeline is in the "normal" sitemap. > > Matthew > > -- > Open Source Group sunShine - Lighting up e:Business > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > Matthew Langham, S&N AG, Klingenderstrasse 5, D-33100 Paderborn > Tel: +49-5251-1581-30 [mlangham@sundn.de - http://www.sundn.de] > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > > > -----Urspr=FCngliche Nachricht----- > Von: Matthew Langham [mailto:mlangham@sundn.de] > Gesendet: Samstag, 28. Juli 2001 15:26 > An: cocoon-dev@xml.apache.org > Betreff: [C2B2] Big Bug in sub-sitemap handling (?) > > > I think I have found a bug in the sub-sitemap handling. Unfortunately (for > me anyway) Carsten is on holiday - so maybe someone else has an idea.. > > I have tested memory consumption using Jprobe. From the results it seems > that pipelines in sub-sitemaps are not released from memory. This > means that > after a request has been processed all the components (!) are > still held in > memory. This means that repetitive requests cause the memory to fill up. > > I tested the same pipeline in the normal sitemap and everything was fine. > > Can someone confirm this - and maybe offer a solution (apart from > not having > sub-sitemaps)? > > Matthew > > -- > Open Source Group sunShine - Lighting up e:Business > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > Matthew Langham, S&N AG, Klingenderstrasse 5, D-33100 Paderborn > Tel: +49-5251-1581-30 [mlangham@sundn.de - http://www.sundn.de] > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > > <<<<<<<<<<<<<<<<<<<<<<<< > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org > For additional commands, email: cocoon-dev-help@xml.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org For additional commands, email: cocoon-dev-help@xml.apache.org