On Mon, 08 Oct 2001, Ian Clarke wrote:

> It seems that people attempting large insertions into the not-so-big 0.4
> network are causing problems.  Please refrain from attempting large
> insertions until the network is larger and the load is better
> distributed.  Also, if you must attempt an insertion, please do so on a
> mature node, rather than a new node whose datastore is full of the
> seedNodes nodes - which are getting slashdotted.
> 
my seed node is quite busy.  I have plenty of bandwidth, but not so much
in the way of CPU or datastore size.

> I would rather we didn't have to implement code to specifically deal
> with this irritation (such as placing limits on the incoming traffic
> permitted from any one node in a given period), as it will only be
> relevent in these early stages.
> 
very agreed.

> Inserting large files is pointless anyway until we get redundant
> splitfiles (see GJ's freesite for why).
> 
Weren't you one of the people insisting that we wouldn't need redundancy
for splitfiles?  Why does having something published in freenet change
your mind about this topic?

> This is also good motivation for us to implement the seed.ref extraction
> code in FCP that I mentioned recently to reduce the reliance on a small
> number of seed nodes (and address Oskar's security concerns).
> 
> Ian.

That's still going to be a centralized approach.  As much as I like the
idea as a stopgap measure, I'd prefer to have it only give one address
per request or something like that to severly limit any usefulness in
getting a list of all nodes in the network.

Thelema
-- 
E-mail: [EMAIL PROTECTED]        If you love something, set it free.
GPG 1536g/B9C5D1F7 fpr:075A A3F7 F70B 1397 345D  A67E 70AA 820B A806 F95D

Attachment: msg02836/pgp00000.pgp
Description: PGP signature

Reply via email to