[volunteergrid2-l] Recommended settings

Billy Earney billy.earney at gmail.com
Wed Jun 29 12:36:17 PDT 2011


"Also automatically generating the config options does not seem too sensible
to me, since the should change dynamically at runtime (especially in bigger
and more volatile grids)."

 

What I meant was that we wouldn't set "raw" numbers in the config file.  But
if we set in the config file something  like N=90%, which would mean to set
N=T*90% (where T is the total number of nodes available at time of file
upload), then H could equal some minimal number of nodes necessary for an
upload, calculated from R (R for reliability %).

 

If we assume that node availability is 95%, then what would H have to be to
have a reliability of R?  There's probably a formula for this in the tahoe
api.

 

My reliability threshold may be 99% for some files (willing to loose
sometimes), but for other files it could be (99.999%).  Which brings up
another topic of allowing these to be entered from the command line when
uploading files, since different files could have different R's.   Just my
$0.02. J

 

I guess this is a little off topic and should be addressed by the regular
tahoe listserv and not the volunteer2 listserv.   But do users of the
volunteer2 grid have suggestions, comments, (or am I just crazy! ) J

 

Billy

 

 

From: volunteergrid2-l-bounces at tahoe-lafs.org
[mailto:volunteergrid2-l-bounces at tahoe-lafs.org] On Behalf Of Shawn Willden
Sent: Wednesday, June 29, 2011 2:11 PM
To: a small group of people sharing storage resources with Tahoe-LAFS
Subject: Re: [volunteergrid2-l] Recommended settings

 

On Wed, Jun 29, 2011 at 7:41 AM, Marco Tedaldi <marco.tedaldi at gmail.com>
wrote:

Just one thought on the settings:

Setting N to the number of nodes might not be the best idea anyway. At least
if this node also provides storage. I think, there is not much sense in
putting my data on my own node.

 

Setting N to T-1 (I'm going to use T to mean "total number of nodes" will
not prevent your node from getting a share.  Server selection is basically
randomized so if all nodes in the grid are performing well, and T=10 and
N=9, your node will get one of your shares 90% of the time.  If some of the
nodes in the grid are slow to respond, your node will get one of your shares
*more* than 90% of the time.

 

For the backup use case this behavior isn't ideal.  But for a lot of other
use cases it makes a great deal of sense.

 

Setting the values automatically has some benefits. But there would have to
be quite some code behind it. Imagine a network with 50 nodes or more...
Setting "shares.total" to 48 would not make too much sense in my oppinion.

 

Why not?

 

Also automatically generating the config options does not seem too sensible
to me, since the should change dynamically at runtime (especially in bigger
and more volatile grids).

 

Right, my goal would be to calculate the N and K values dynamically, for
every upload.

 

-- 
Shawn

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tahoe-lafs.org/cgi-bin/mailman/private/volunteergrid2-l/attachments/20110629/ec37522d/attachment.html>


More information about the volunteergrid2-l mailing list