On Tue, Aug 29, 2000 at 05:52:52PM +0100, degs wrote: > I'm beginning to suspect my machine here...
It's more likely to be Freenet actually. I mean, you probably would have noticed if your machine where incapable of local connections. The socket dying in the middle of the send like it did before was pretty damn weird though I must admit, I thought maybe Freenet's connection timeout code wasn't making note of the fact that the connection was being written to - but I checked and the code is there, and besides that happens after 3 minutes not 19 seconds (unless you have changed the connectionTimeout setting, which might be easy to confuse with connectTimeout if you meant to change that...) > Recent CVS code changes the behaviour, but I still can't insert anything - > inserts now end with 'failed gracefully' and node log contains many > 'ConnectTimedOut' exceptions. Netstat shows these attempted connections > stuck in SYN_SENT and attempting to contact most of these hosts from another > network shows they really are either dead or not running a node. However > references to these nodes seem to stick in my data store - shouldn't they be > evicted after the first attempt to handshake fails? Even on htl 1 inserts like the on from the log before? I don't know what other nodes are active on the dev Freenet, but it wouldn't surprise me if they are broken. Reference are removed the first time they fail, but only if the node actually manages to send the data somewhere - this is to prevent the whole DataStore from being cleared of references on a single request because the network connection is temporarily down. If your Requests are not making it anywhere, it will therefore not remove the references (this makes sense, if it cleared your datastore of all connection > Anyway I'm going to try this on some different machines to try and find out > what aspect of my setup here is causing problems. > > If this does turn out to be my machine then I apologise for wasting your > time Oskar... There's always something to learn. > > Degs > > > ----- Original Message ----- > From: Oskar Sandberg <md98-osa at nada.kth.se> > To: <freenet-dev at lists.sourceforge.net> > Sent: Tuesday, August 29, 2000 2:30 PM > Subject: Re: [Freenet-dev] Dynamic DNS/NAT not working > > > > On Tue, Aug 29, 2000 at 01:46:55PM +0100, degs wrote: > > > > > > Hi, > > > > > > Has anyone had any success running a node on a private subnet via a > > > forwarded port and using dynamic DNS? I'm still having trouble here... > > > > > > I have included node and client logs from an attempt to freenet_insert > > > > > > freenet_insert is running from a remote machine (at work). node is > running > > > on machine at home behind ISDN router. node has its address set to > > > ceequdee.no-ip.com which does not resolve to the machine the node is > running > > > on, it resolves to the router which has a hole at 19114 pointing at the > > > node. > > > > The error you are getting has nothing to do with the address, you are > > getting an IOException because the connection is being closed while you > > are sending the data: > > > > The node says: > > > PM:Freenet.support.io.PadOnErrorInputStream:-1830150648:Normal:Exception > > > caught: java.net.SocketException: Socket closed padding... > > > Aug 29, 2000 12:22:57 PM:Freenet.Conduit:-2046943736:Minor:0- Exception: > > > Freenet.support.io.DataNotValidIOException: Data not valid, code 1 > thrown > > > during read of Conduit feed > > > > Code 1 means that bad data was discovered. Which is quite correct (sort, > > of I'm thinking it should be code 0, but that is not relevant > > here) because the client is failing to send the data: > > > > > Aug 29, 2000 1:28:09 > PM:Freenet.ConnectionHandler:1802907411:Debug:Finished > > > with connection - closing > > > Aug 29, 2000 1:28:09 > > > > PM:Freenet.client.EventInputStream$TransferedEvent:-282185966:Normal:262144 > > > bytes read. > > > Aug 29, 2000 1:28:09 PM:Freenet.Conduit:-1504039150:Minor:0- Exception: > > > java.io.IOException: Bad file descriptor thrown during write of Conduit > feed > > > Handled: Freenet.client.Client$BInsertInstance$SentInsert at 67f62f13 > > > > It looks like an IOException occurs while reading from the socket, so the > > listening thread procedes to close the socket. What the IOException is I > > don't know since it is not being output to the log. Is this happening at > > the exact same time every time? Does it matter how big the files are? > > > > > > -- > > \oskar > > _______________________________________________ > > Freenet-dev mailing list > > Freenet-dev at lists.sourceforge.net > > http://lists.sourceforge.net/mailman/listinfo/freenet-dev > > > > _______________________________________________ > Freenet-dev mailing list > Freenet-dev at lists.sourceforge.net > http://lists.sourceforge.net/mailman/listinfo/freenet-dev > -- \oskar _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
