On Monday 11 November 2002 20:56, Mike wrote: > A bit more info, > This was with 0.5.0.6 and it may have occured on a GMT date switch. > Considering that it was a CHK FEC splitfile pointed to by a edition > based freesite I don't think that it should have affected it but it > might be an avenue to check. I wasn't around to actually see the *exact* > time it failed so the GMT thing might not be relavant. > > Mike >
The dbr should be irrelevant. Once the downloader is running it already has all the SplitFile metadata. >> java.io.IOException: Couldn't read one or more segments from Freenet. I don't think this is an error. It just means that fproxy couldn't get enough blocks to decode the segment so it gave up on the splitfile. I is admittely bad UI. >> Request for an unknown, finished or expired context. The download servlet only keeps information about for a limited time after a request finishes (30 minutes I think). Fproxy is dumb about stopping refreshing of the bottom pane once the request stops. You only see the final status message until the request context is freed, then you will get the above message. > Is there a upper limit to the size of a splitfile based on freenet only > waiting so long for you to download all the chunks? There is some code that will abort a request if none of the download threads reports any status information for a long time (30 minutes). The rationale was to give up when your node is so busy that it can't answer requests. You might be able to make it trip on a healthy node if you used ridiculously high htl values. > > vvvvvv original message vvvvvvvvv > Got this message on console in attempting to download a large splitfile: > > Starting redundant download of segment 0. Need 128 of 192 blocks. > java.io.IOException: Couldn't read one or more segments from Freenet. > at freenet.support.BucketSequence.handleError(BucketSequence.java:233) > at freenet.support.BucketSequence.readable(BucketSequence.java:247) > at freenet.support.BucketSequence.read(BucketSequence.java:289) > at freenet.support.BucketSequence.access$1(BucketSequence.java) > at > freenet.support.BucketSequence$BucketSequenceInputStream.read(BucketSequenc >e.java:354) at > freenet.client.http.FproxyServlet.copy(FproxyServlet.java:857) at > freenet.client.http.SplitFileRequestContext.sendData(SplitFileRequestContex >t.java:929) at > freenet.client.http.SplitFileRequestContext.renderPage(SplitFileRequestCont >ext.java:330) at > freenet.client.http.SplitFileRequestContext.handle(SplitFileRequestContext. >java:103) > > at > freenet.client.http.FproxyServlet.handleContexts(FproxyServlet.java:380) > at freenet.client.http.FproxyServlet.doGet(FproxyServlet.java:412) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:740) > at freenet.client.http.FproxyServlet.service(FproxyServlet.java:312) > at javax.servlet.http.HttpServlet.service(HttpServlet.java:853) > at > freenet.interfaces.servlet.ServletContainer.handle(ServletContainer.java:62 >) at > freenet.interfaces.LocalInterface$ConnectionShell.run(LocalInterface.java:1 >64) at freenet.thread.QThreadFactory$QThread.run(QThreadFactory.java:212) > This query failed to deallocate ConnectionHandler$ReceiveInputStream > allocated for AnnouncementExecute, 0 of 41 bytes done. > > And a strange message on the splitfile browser screen: > > > Request for an unknown, finished or expired context. > > > Is there a upper limit to the size of a splitfile based on freenet only > waiting so long for you to download all the chunks? > > Mike > > > > _______________________________________________ > devl mailing list > devl at freenetproject.org > http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl _______________________________________________ devl mailing list devl at freenetproject.org http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl
