[volunteergrid2-l] Should we increase the node minimum capacity?

Fri Nov 4 22:31:30 UTC 2011

Good questions... and made even better by the fact that kicking someone off
the grid is actually not very easy to do.  Migrating the data isn't so much
of a problem... it's just like having a node (or some nodes) die, so the
normal repair process will fix things up.  But kicking someone off a grid
basically requires starting a new grid and moving all the storage nodes
over to it.  More precisely, you have to set up a new introducer, publish
the FURL to everyone (except those being ejected) and then change all of
the nodes to use the new introducer.  After first giving the ejectees a
reasonable grace period to download their data, of course.

However, if someone were to announce that they refuse to upgrade, I would
hope that many of us would vote against the policy change rather than
create a situation where we have to exercise that "nuclear option".  I
certainly would.

On Fri, Nov 4, 2011 at 4:11 PM, Brad Rupp <bradrupp at gmail.com> wrote:

> Six months seems reasonable to me too.  The CIVS process is also fine with
> me.
>
> One question I have.  Say we decide a 3/4 majority wins.  What happens to
> the 1/4 that voted against something?  In the case of hard drive sizes, we
> certainly can't force them to upgrade to a bigger hard drive.  They might
> not have funds to do so.  At that point what happens?  Are  they kicked off
> the grid and we have to migrate their data?  We certainly need to explore
> what will happen to those with a dissenting vote.
>
> I realize that not all policy decisions will be so rash.  But hard drive
> sizing could get that way.
>
> Brad
>
>
> On 11/4/2011 3:37 PM, Shawn Willden wrote:
>
>> Six months seems perfectly reasonable to me.  I'm not sure what it
>> means to give a "minimum" of six months, though... perhaps what you
>> mean would be more clearly expressed as:
>>
>> "In the event that the VG2 community elects to increase the minimum
>> node capacity requirement, the community will also choose a timeframe
>> within which all existing members must meet the new requirement.  The
>> timeframe will be no sooner than six months after the requirement
>> change is approved.  New members and new nodes will be required to
>> commit to the new guideline upon joining."
>>
>> If that's what you meant, that the six months is a binding minimum on
>> the community, that we cannot make a change faster than that, I wonder
>> if it should be a little tighter; maybe three months.  Keep in mind
>> that although our 2:1 ratio does help keep us from getting to the
>> effectively full state, it doesn't absolutely prevent it.
>>
>> This also points out another question... how do we know when the
>> community has made a decision to change a policy?  It seems like we
>> need to define a voting process, and a voting standard.  Do we want to
>> formally state that all policy changes must be by consensus -- by
>> which I mean "general accord", or 100% approval?  A simple majority
>> seems to me not to be a sufficient basis on which to change policies,
>> but 100% approval is problematic.  A 2/3 or 3/4 supermajority?  That
>> seems like a good approach to me.
>>
>> As for process, there's also the question of whether we want balloting
>> to be secret or public, and how to go about it.  I started to type out
>> a proposal and then thought... someone has to have created a nice,
>> free on-line service for this, so I searched and found
>> http://www.cs.cornell.edu/**andru/civs.html<http://www.cs.cornell.edu/andru/civs.html>.
>>  It seems perfect, to me.
>>
>> I propose we use CIVS for policy decisions.  We can have a list
>> discussion to get general consensus on the question and the choices,
>> and then we'll use the service, giving one vote to each NODE.  The
>> list of voter e-mail addresses will be from the list of active,
>> operational nodes.  Operators who have multiple nodes will have to
>> provide the vote supervisor with additional addresses they can use to
>> submit their other votes if they want to exercise their multi-voting
>> prerogative.
>>
>> Just as a test, I've created a CIVS vote on this node increase
>> question.  Everyone who receives an invitation should please vote, but
>> we won't consider the result binding.  It's mostly just a test, and I
>> figure it might also be informative.
>>
>> On Fri, Nov 4, 2011 at 3:01 PM, Jody Harris<jharris at harrisdev.com>
>>  wrote:
>>
>>> A policy proposal:
>>> "In the event that the VG2 community elects to increase the minimum node
>>> requirement, all members will be given a minimum of 6 months to comply
>>> with
>>> the new guideline. New members and new nodes will be required to commit
>>> to
>>> the new guideline upon joining."
>>> I guess this would go in as "4b" in the policies
>>> (http://bigpig.org/twiki/bin/**view/Main/**VolunteerGrid2Policies<http://bigpig.org/twiki/bin/view/Main/VolunteerGrid2Policies>
>>> ).
>>> Thoughts?
>>> 6 months?
>>> jody
>>> ----
>>> Ph. 575-208-4567
>>> - Think carefully.
>>>
>>>
>>> On Fri, Nov 4, 2011 at 3:28 PM, Shawn Willden<shawn at willden.org>  wrote:
>>>
>>>>
>>>> So far all of the responses have been pretty positive.
>>>>
>>>> So... is there anyone who objects to increasing the minimum node side
>>>> from 500 GB to 1 TB (and the maximum size from 1 TB to 2 TB)?  I don't
>>>> want anyone to feel like they're being railroaded.
>>>>
>>>> On Fri, Nov 4, 2011 at 2:20 PM, Marco Tedaldi<marco.tedaldi at gmail.**com<marco.tedaldi at gmail.com>
>>>> >
>>>> wrote:
>>>>
>>>>> Hello everyone
>>>>>
>>>>> Okay... I've just upgraded my server to 1TB about a year ago... Seems
>>>>> to
>>>>> be time to do it again.
>>>>>
>>>>> Good to have a good reason to do so.
>>>>>
>>>>> best
>>>>>
>>>>> Marco
>>>>>
>>>>> Am 03.11.2011 21:30, schrieb Christoph Langguth:
>>>>>
>>>>>> Hi all,
>>>>>>
>>>>>> this is just to let you know that we just ordered the new 2TB disk
>>>>>> today. I hope to be joining the grid within the next 2 weeks or so.
>>>>>>
>>>>>> In any case, I'm perfecly fine with increasing the minimum -- the new
>>>>>> disk will be used for VG2 and local backups exclusively, so 1TB is no
>>>>>> problem at all.
>>>>>>
>>>>>> Cheers
>>>>>> Chris
>>>>>>
>>>>>>
>>>>>>
>>>>>> ______________________________**_________________
>>>>>> volunteergrid2-l mailing list
>>>>>> volunteergrid2-l at tahoe-lafs.**org <volunteergrid2-l at tahoe-lafs.org>
>>>>>> http://tahoe-lafs.org/cgi-bin/**mailman/listinfo/**volunteergrid2-l<http://tahoe-lafs.org/cgi-bin/mailman/listinfo/volunteergrid2-l>
>>>>>> http://bigpig.org/twiki/bin/**view/Main/WebHome<http://bigpig.org/twiki/bin/view/Main/WebHome>
>>>>>>
>>>>>
>>>>> ______________________________**_________________
>>>>> volunteergrid2-l mailing list
>>>>> volunteergrid2-l at tahoe-lafs.**org <volunteergrid2-l at tahoe-lafs.org>
>>>>> http://tahoe-lafs.org/cgi-bin/**mailman/listinfo/**volunteergrid2-l<http://tahoe-lafs.org/cgi-bin/mailman/listinfo/volunteergrid2-l>
>>>>> http://bigpig.org/twiki/bin/**view/Main/WebHome<http://bigpig.org/twiki/bin/view/Main/WebHome>
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Shawn
>>>> ______________________________**_________________
>>>> volunteergrid2-l mailing list
>>>> volunteergrid2-l at tahoe-lafs.**org <volunteergrid2-l at tahoe-lafs.org>
>>>> http://tahoe-lafs.org/cgi-bin/**mailman/listinfo/**volunteergrid2-l<http://tahoe-lafs.org/cgi-bin/mailman/listinfo/volunteergrid2-l>
>>>> http://bigpig.org/twiki/bin/**view/Main/WebHome<http://bigpig.org/twiki/bin/view/Main/WebHome>
>>>>
>>>
>>>
>>> ______________________________**_________________
>>> volunteergrid2-l mailing list
>>> volunteergrid2-l at tahoe-lafs.**org <volunteergrid2-l at tahoe-lafs.org>
>>> http://tahoe-lafs.org/cgi-bin/**mailman/listinfo/**volunteergrid2-l<http://tahoe-lafs.org/cgi-bin/mailman/listinfo/volunteergrid2-l>
>>> http://bigpig.org/twiki/bin/**view/Main/WebHome<http://bigpig.org/twiki/bin/view/Main/WebHome>
>>>
>>>
>>
>>
>>  ______________________________**_________________
> volunteergrid2-l mailing list
> volunteergrid2-l at tahoe-lafs.**org <volunteergrid2-l at tahoe-lafs.org>
> http://tahoe-lafs.org/cgi-bin/**mailman/listinfo/**volunteergrid2-l<http://tahoe-lafs.org/cgi-bin/mailman/listinfo/volunteergrid2-l>
> http://bigpig.org/twiki/bin/**view/Main/WebHome<http://bigpig.org/twiki/bin/view/Main/WebHome>
>

-- 
Shawn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tahoe-lafs.org/cgi-bin/mailman/private/volunteergrid2-l/attachments/20111104/efd5738a/attachment-0001.html>