[tahoe-dev] UnhappinessError on volunteer grid

[Kevan Carstensen]

At 2010-09-16 00:10 (+0200), slush wrote:

> Hello,
> 
> I'm using Tahoe's VolunteerGrid for my personal backups.  I set up
> backups and left it for few weeks ago without special attention. But I
> tried to upload more files last week and found UnhappinessError:
> 
> <class 'allmydata.interfaces.UploadUnhappinessError'>: shares could be
> placed on only 4 server(s) such that any 3 of them have enough shares
> to recover the file, but we were asked to place shares on at least 6
> such servers. (placed all 12 shares, want to place shares on at least
> 6 servers such that any 3 of them have enough shares to recover the
> file, sent 16 queries to 12 peers, 8 queries placed some shares, 8
> placed none (of which 8 placed none due to the server being full and 0
> placed none due to an error))
> 
> I tried it today again and got again the same error. On Tahoe Welcome
> page are currently online 17 servers, so I'm surprised that all of
> them are full or readonly? I don't believe VolunteerGrid is full as I
> provide two servers with capacity over 2TB and consume only around
> 60GB of space...?

Can you check to see which version of Tahoe-LAFS the storage servers in
the VolunteerGrid report that they are using? You can see this
information on the page opened by the 'tahoe webopen' command, in the
'Version' column. Also, which version of Tahoe-LAFS are you using?

When we first released 1.7.0, we discovered that there were some bugs in
the way that Tahoe-LAFS (specifically, the immutable uploader and the
storage server) handled incomplete immutable uploads [1]. One of the
fixes in 1.7.1 addressed these issues. If you find that a large number
of storage servers are running versions of Tahoe-LAFS older than 1.7.1,
it may explain this behavior, since we first discovered the problem when
another user reported upload behavior much like yours [2].

Can you say more about how you are backing up your files? Are you using
the 'tahoe backup' command, or some other technique? If it turns out that
most servers are running 1.7.1 or later, that information might help us
to track down the problem.

[1] http://tahoe-lafs.org/trac/tahoe-lafs/ticket/1117
[2] http://tahoe-lafs.org/pipermail/tahoe-dev/2010-July/004656.html

Thanks,
-- 
Kevan Carstensen | <kevan at isnotajoke.com>