Cannot upload files to Tahoe anymore

[Jean-Paul Calderone]

On Wed, Oct 2, 2019 at 11:47 AM Gwenaëlle Fournier <
gwenaelle.fournier at gmail.com> wrote:

> Hi tahoe dev list,
>

Hiya!


>
> I come to you with a problem that occurred while switching from tahoe
> 1.11.0 to 1.13.0.
>
> We have an API to store files using tahoe.
> For now we have been using tahoe v1.11.0 but we are thinking of migrating
> to tahoe v1.13.0 and I tested doing that on my local environment.
>
> I have been able to v1.13.0 on my environment:
> - git clone from master branch
> - virtualenv venv
> - ./venv/Scripts/pip install --editable .[test]
> - create client, node and introducer and configure tahoe.cfg files
>

What configuration did you put into your tahoe.cfg files - particularly for
the client node?  If you can share the whole file, that'd be great.  I
specifically wonder what `shares.*` are set to.


> Now I can run tahoe and my Tahoe-LAFS WUI shows everything green with my
> introducer connected and two nodes (client and node).
> Everything seems fine here.
>
> From our API, I am still able to access the files I had previously
> uploaded while still in v1.11.0.
>
> But, and here is my problem, I cannot upload files anymore to tahoe.
> *Request* (performed by our API):
> PUT =>
> http://127.0.0.1:3456/uri/URI:DIR2:qxrozrgcqcsu5y6pbkkmg6v26m:b4vjn7vlqmptnfptfebdq5qnzehruvo5v5oblw7j752gayfsfrpq/636_test.jpg
>
> *Response*:
> status: 410
> body:  UnrecoverableFileError: the directory (or mutable file) could not
> be retrieved, because there were insufficient good shares. This might
> indicate that no servers were connected, insufficient servers were
> connected, the URI was corrupt, or that shares have been lost due to server
> departure, hard drive failure, or disk corruption. You should perform a
> filecheck on this object to learn more.
>
> Tahoe exact version: *1.13.0.post1207*
>
> There are *no log files* in client logs, node logs or introducer logs and
> nothing also in twistd.log.
>
> I have been able to track the error down in tahoe's code until
> *_get_version_from_servermap*() function in
> *src/allmydata/mutable/filenode.py* but I don't understand it (I am very
> new to tahoe).
>
> An additional problem is that if I switch back to version v1.11.0, I still
> have the issue. I surely broke something somewhere...
>

How did you switch back to v1.11.0?  Did you just use pip to install that
version of Tahoe-LAFS into the same env you were using?  I don't know if
downgrading is particularly reliable - but even if it worked, if there is a
problem in tahoe.cfg, v1.11.0 would inherit it if you did things this way.

Regarding logs, you might want to take a look at `flogtool` which can
extract information from the Foolscap incident logs, which is where
Tahoe-LAFS tries to send most information.

Jean-Paul


>
> Could you please help me?
>
> Thanks a lot in advance,
>
> Gwenaelle
> _______________________________________________
> tahoe-dev mailing list
> tahoe-dev at tahoe-lafs.org
> https://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tahoe-lafs.org/pipermail/tahoe-dev/attachments/20191002/6af4c6d5/attachment.html>