On Sunday 19 Aug 2012 15:09:32 Dennis Nezic wrote:
> On Wed, 15 Aug 2012 14:29:15 +0100, Matthew Toseland wrote:
> > On Wednesday 15 Aug 2012 05:05:44 Dennis Nezic wrote:
> > > On Tue, 14 Aug 2012 23:50:03 -0400, Steve Dougherty wrote:
> > > > On 08/14/2012 10:52 PM, Dennis Nezic wrote:
> > > > > On Tue, 14 Aug 2012 11:40:49 -0400, Steve Dougherty wrote:
> > > > >> On 08/14/2012 08:56 AM, Dennis Nezic wrote:
> > > > >>> On Fri, 10 Aug 2012 15:47:35 -0400, Steve Dougherty wrote:
> > > > >>>> Any idea what may have changed before this problem started
> > > > >>>> happening? What version of ant are you using to build?
> > > > >>>
> > > > >>> That's what I'm pulling my hair trying to find out! I only
> > > > >>> upgraded a few system packages on my machine, like glib,
> > > > >>> libffi, ... -- and nothing much else! But how can a seemingly
> > > > >>> unrelated system package have such a huge effect in a Java
> > > > >>> environment??
> > > > >>>
> > > > >>> I haven't changed my jdk (although recently I did test with
> > > > >>> others, sun-jdk and icedtea-bin), I haven't changed my ant, I
> > > > >>> haven't changed my db4o's, I recompiled all of Freenet's
> > > > >>> (direct) dependencies. This is incredibly frustrating. I mean,
> > > > >>> 1407 WAS working before, and now it's not :S.
> > > > >>>
> > > > >>> Here are the types of AbstractMethodError's that arise:
> > > > >>>
> > > > >>>
> > > > >>> java.lang.AbstractMethodError:
> > > > >>> com.db4o.internal.query.ObjectSetFacade.iterator
> > > > >>> ()Ljava/util/Iterator;
> > > > >>>
> > > > >>>
> > > > >> at freenet.client.async.ClientRequestScheduler.loadKeyListeners
> > > > >>> (ClientRequestScheduler.java:114) [code]
> > > > >>> ObjectSet<HasKeyListener> results = Db4oBugs.query (container,
> > > > >>> HasKeyListener.class); for (HasKeyListener l : results) {
> > > > >>> <--- ** 114 [/code]
> > > > >>>
> > > > >>>
> > > > >>> java.lang.AbstractMethodError:
> > > > >>> com.db4o.internal.query.ObjectSetFacade.iterator
> > > > >>> ()Ljava/util/Iterator;
> > > > >>>
> > > > >>>
> > > > >> at
> > > > >> freenet.client.async.PersistentStatsPutter.restorePreviousData
> > > > >>> (PersistentStatsPutter.java:44) [code]
> > > > >>> ObjectSet<BandwidthStatsContainer> BSCresult =
> > > > >>> container.query (BandwidthStatsContainer.class); for
> > > > >>> (BandwidthStatsContainer candidate : BSCresult) { <--- **
> > > > >>> 44 [/code]
> > > > >>
> > > > >> So what did you do to establish that your database is not
> > > > >> corrupted? These are internal db4o errors.
> > > > >
> > > > > Hmm. So it does seem as though my datastore was corrupted.
> > > > > Deleting it, and making my node create a new one got me up and
> > > > > running again. Thanks Steve!
> > > > >
> > > > > The odd thing is a bunch of us[1] have been getting this
> > > > > error / corruption since about the time build1409 came out. I
> > > > > sure hope it's just a freak coincidence :S.
> > > > >
> > > > > [1] https://bugs.gentoo.org/show_bug.cgi?id=429716
> > > >
> > > > Your _datastore_ was corrupted, not your node.db4o(.crypt)?! What
> > > > exactly did you do to get your node running again?
> > >
> > > Correct. The very first thing I tried (which usually worked before
> > > with db4o errors), was to delete my node.db4o.crypt file. That had
> > > no effect this time. Instead, deleting datastore/*, or starting
> > > with a fresh installation, with the wizard, got the node running.
> > > Drastic, I know. (Although, Tommy in the bug report mentioned he
> > > was able to reproduce the error even with a clean wizard
> > > installation :S.)
> >
> > That is not consistent with the error message you gave, which is
> > clearly a db4o problem.
>
> Are you sure about that? Is there no way a "corrupt" datastore could
> corrupt the db4o file? What do you suppose created that db4o Abstract
> Method Error? (Considering the fact that deleting it, and letting
> freenet make a new one still caused it.)No, this would definitely not be caused by a problem with the datastore, which is a very simple structure that does not store any code or complex objects. Whereas db4o could well cause such problems - either spontaneously, on defrag, on doing uploads, or ... Well we have lots of reports and we're not quite sure what to do about it yet. :(
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ Support mailing list Support@freenetproject.org http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://emu.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:support-requ...@freenetproject.org?subject=unsubscribe
