[volunteergrid2-l] Very soon after uploading, checks failing.

[Aaron Drew]

I've spent the past few weeks repairing my unhealthy data. (My uplink
sucks.) Just when I get to the end, I run:

 $ tahoe deep-check --repair --add-lease --verbose ...

And I see more unhealthy files that I just finished repairing.

My completely unconfirmed hypothesis is that some people have
expire_override_lease_duration set to something tiny. Would that explain
this? If so, since others have had similar issues, would a rule for the
volunteer net be possible setting a minimum lease period of (say) a month?

On 25 November 2012 16:39, David-Sarah Hopwood <david-sarah at jacaranda.org>wrote:

> On 23/11/12 11:06, Aaron Drew wrote:
> > This is very worrying.
> >
> > I have spent about 2 weeks uploading 60GB of my most important data to
> tahoe (using
> > duplicity for backups) and as soon as I'm finished, I do a check and get
> this output:
> >
> > $ tahoe deep-check riot-backup:
> ...
> > done: 2147 objects checked, 592 healthy, 1555 unhealthy
>
> "Unhealthy" means that not *all* shares were uploaded. So it would only
> take one server
> to be intermittently reachable to get a result like this.
>
> --
> David-Sarah Hopwood ⚥
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://tahoe-lafs.org/cgi-bin/mailman/private/volunteergrid2-l/attachments/20121209/a4a906bd/attachment.html>