[Freedombox-discuss] How to deal with uptime and connectivity issues

Matt Willsher matt at monki.org.uk
Tue Mar 1 20:12:49 UTC 2011


On 1 March 2011 18:13, Michael Blizek <
michi1 at michaelblizek.twilightparadox.com> wrote:

> Hi!
>
> On 04:52 Tue 01 Mar     , Felipe Sanches wrote:
> > SaaS users usually benefit from high availability provided by
> professional
> > datacenters. Data is often stored redundantly.
>
> In if the datacenter operator looses motivation to host you, you will spend
> days to find and set up an alternative. That is, if you have made your own
> backup, because backups in the datacenter will be unaccessable...
>

The future of cloud, if it's to stay around, is standardisation which will
allow seamless migration between providers.  Cloud as it is now is still
pretty young.


> You might get 99,999% availability for 10 years, but you you will then go
> down
> for 10 days, your average availability is only 99,7%.




> > How will we deal with those issues in the scenario of a single plug per
> user
> > tied to a power outlet? Are we expecting people to have multiple
> > freedomboxes kept in sync? I foresee people unplugging someone else's
> boxes
> > in order to use the power outlet.
>
> I do not think we should worry too much about these issues.
>

I don't think we do for 0.1, or even 1.0 but users are beginning to expect
reliability in their services.

Much of the benefits the that were or still are the mainstay of the
enterprise are available to people free it they can be
correctly architectured. I think the FB should leverage redundancies where
possible. Given  the price point target of 29USD its not infeasible to
provide very high uptime through a cloud like structure to the FBs. A UPS,
two FBs and 99.99999% uptime can be yours ;)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/freedombox-discuss/attachments/20110301/8ec0f74e/attachment.htm>


More information about the Freedombox-discuss mailing list