<div dir="ltr"><div>1</div>yes there was an helper working<div><br></div><div>this is necessary in our design</div><div>as the nodes are connected by low speed adsl and the helper is on a high speed internet connection.<br>
</div><div><br></div><div><br></div><div>actually it was not possible to backup more that 2-5 gByte of data at same time. directly on the global network</div><div>we always go to a kind of error or infinite loop.</div><div>
<br></div><div><br></div><div><br></div><div>2</div><div> we want to backup at least 100 GByte of existing data per node plus daily history</div><div>and i never got anything working fine</div><div><br></div><div><br></div>
<div>i managed to make a workaround</div><div>this is to make a tahoe backup on two nodes on a local network with one local introducer.</div><div>that way the shares are create and the backup folder is filled with the filestructure</div>
<div>then to change configuration of the nodes with the global introducer</div><div>then to repair the folder so that the shares get a better repartition on the network</div><div><br></div><div><br></div><div>that way i can manage to get the first sync to be done fast </div>
<div>and then the daily update with 'tahoe backup' can be done easily</div><div><br></div><div><br></div><div>the two nodes can be in two machines or on the same machine</div><div><br></div><div><br></div><div>this is mainly to circumvent the fact that file structure is only created at the end of the "tahoe backup" and not file by file</div>
<div>like a more common "rsync".</div><div>tested rsync over tahoe/sshfs and results ware very bad. after hours rsync was still creating the folders with no file uploaded</div><div><br></div><div><br></div><div>
<br></div><div><br></div><div><br></div><div>cordialement,</div><div><br></div><div>xavier lebrun</div><div><br></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-07-29 15:03 GMT+02:00 Daira Hopwood <span dir="ltr"><<a href="mailto:daira@jacaranda.org" target="_blank">daira@jacaranda.org</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="">On 11/07/14 15:32, freelab xavier wrote:<br>
> the error messages are not that clear<br>
<br>
</div>Yes, you're right. Sorry about that.<br>
<div class=""><br>
> what shall I understand from those errors ?<br>
<br>
</div>The storage server reported an exception that appears to be a bug in code related to<br>
the upload helper. I haven't seen this particular error before. Is a helper enabled?<br>
<br>
In any case, I have filed this as<br>
<<a href="https://tahoe-lafs.org/trac/tahoe-lafs/ticket/2267" target="_blank">https://tahoe-lafs.org/trac/tahoe-lafs/ticket/2267</a>><br>
<br>
[...]<br>
<div class="HOEnZb"><div class="h5">> File \"/usr/lib/pymodules/python2.7/allmydata/immutable/offloaded.py\", line 203, in<br>
> remote_upload<br>
> self._reader.add_reader(reader)<br>
> exceptions.AttributeError: 'CHKUploadHelper' object has no attribute '_reader'<br>
> ]'><br>
> "<br>
<br>
</div></div><span class="HOEnZb"><font color="#888888">--<br>
Daira Hopwood ⚥<br>
<br>
</font></span></blockquote></div><br></div>