[tahoe-dev] split brain? how handled in tahoe -- docs?

Michael Rogers michael at briarproject.org
Wed Aug 8 08:28:25 UTC 2012


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/08/12 08:57, Tony Arcieri wrote:
> These are the only two options. For those who desire
> "reliability", these are the only buckets that reliability can be
> segmented into. So far I have not heard *any* good arguments
> towards strong consistency.

Hi Tony,

Thanks for explaining the CAP theorem. I believe erpo41 made a
suggestion that would switch Tahoe's behaviour from availability to
consistency: make H (also known as shares.happy) greater than half the
number of storage nodes, so writes can only succeed in at most one
component of a partitioned network. Do you think that would work?

Cheers,
Michael
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJQIiMpAAoJEBEET9GfxSfMLJMIAKiGmb+nfpur8tads5dFKwqF
VRcXlVv2XmXOmR2qbpK7TmhfQ+1MPkUjEMSq/Ha8xC5n0Dd6fC4uswSJHX5gWZrU
WkBpPXwdOvf7uf7EfZzUJXG7zJaEoa9H/zNOxY7rSoU97m+YAtOkr3c4ZOZrmeq3
vM6x1MrhJjg5XZqnsOxVzXvTKL6LTcHdzQQDQtItXXF+UHKgzPDZb7DiKlt+Towe
OfBmIcjaArssImjI6p/QLIZyTRmX4vglj2/ht1MgGWVd/7mFmRZ36ZsjOVOSbjyP
81f6lGfqc3U1XZNUWj1MMVl50T9M2PYI8Nl0625kRk9/SKh4E1GaERYzPrYFB68=
=RaI3
-----END PGP SIGNATURE-----



More information about the tahoe-dev mailing list