James,
I've looked at the log. It looks like the code is trying to submit at
version 15 and the WaveletContainer is expecting version 12 on a local
wavelet. I guess the first thing to figure out is what version we
actually should be submitting at in your case. There are two places
to look within mongoDB. First, start up the local mongo client with
mongo.exe (server uses mongod.exe).
at the command prompt type
> use mongowave
you should see message "switched to db mongowave"
now there are two places to look. The first is the indexwave data.
This will give you the index wave info per participant/per wave. Take
a look at the metadata (__md) and you should see the end_version.
This is the last version mongo has for that wave. It should match for
all participants in that wave. The next thing to look at is the delta
information stored about that wave. The metadata contains the
start_version and end_version of the delta. The end_version should
match with what is stored in the index wave. See if these values are
at version 12 or 15. Do this with the following commands, which use a
regex for the key to find offending wave (I use \Q and \E to quote the
text so I don't have to escape the + or anything else that may pop up
in the wave id):
> db.waveserver.find({"__key":/.+\Qw+ETJEGOtGtkAu\E/, "__md.type":"indexwave"});
You should see a record for every participant on that wave and they
should all have the same end_version
> db.waveserver.find({"__key":/.+\Qw+ETJEGOtGtkAu\E/, "__md.type":"delta"});
You should now see a record for each delta with start_version and
end_version in the metadata. The data is base64 encoded so you won't
be able to make any sense out of that, however, if the last thing you
type before you restarted is there after the restart the end_version
here is probably representative of the correct one.
Let me know if the index wave is in sync with the deltas. Also, it
would be helpful to know the scenario by which this wave was created.
Does it fail on the simple case of say starting a local wave, typing
hello world, restarting the server, opening that wave, and updating
text. If that works, we can work our way up by adding a participant,
operating over federation, etc until we see where the breakdown
occurs.
Thanks for testing it out!
R,
Anthony
On Mar 10, 6:26 am, James Purser <[email protected]> wrote:
> Okay, I thought I would have a go at getting the MongoDB patchset up and
> running. I've downloaded the patch and a fresh copy of the mongodb jar.
>
> It builds without complaint (once I'd discovered that applying the patch
> hadn't put the mongo jar where it should have ;)).
>
> Starting up the server for the first time you will have to wait while
> FedOne builds the MongoDB indexy thing (I have no experience with
> MongoDB so excuse the butchering of the terminology). Once the index had
> built I created a new wave, added my wavesandbox user and made some
> changes. I then killed the server and restarted it. Yay my wave was
> there!
>
> However, when I went to continue the "discussion" I hit issues. I've
> included the log from a single simple session, as you can see I am
> hitting hash mismatch issues.
>
> I'm still testing so I'll post what else I find
> --
> James Purserhttp://wavingtheshiny.collaborynth.com.au
> Wave Addresses:
> [email protected] (wave.google.com)
> [email protected] (wavesandbox.com)
> [email protected] (collaborynth.com.au FedOne Server)
> Skype: purserj1977
> GTalk: [email protected]
>
> wave.log
> 6KViewDownload
--
You received this message because you are subscribed to the Google Groups "Wave
Protocol" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/wave-protocol?hl=en.
