Peter Rabbitson ribasushi

mst_20151207.log:[16:42:22] <mst> aight. what stuff are you going to want me to take on?
mst_20151207.log:[18:06:39] <ribasushi> um... we talked about this - I am not doing an orderly handover
mst_20151207.log:[18:06:45] <ribasushi> so.. nothing? :)
mst_20151207.log:[18:07:03] <ribasushi> am going to be @ LPW, can talk more then if you insist
mst_20151208.log:[15:54:23] <mst> you said on the blog post you'll be transferring maint bits. I'm not asking for orderly, just a heads up as to which things I'm going to wake up on boxing day owning, if anyway
mst_20151208.log:[16:39:12] <ribasushi> that bridge will be burning when I get to it
mst_20151208.log:[16:39:12] <ribasushi> don't worry - I will not land you in a Scwern-like situation
mst_20151208.log:[16:39:12] <ribasushi> but that's about it
mst_20151208.log:[16:44:55] <mst> so, basically "giving warning would stop it being the not-orderly thing you have in mind, but while you're trying to make a splash, you're going to do so without causing carnage" or so

Note!!!

This is a reverse-spec, of a currently existing, working production pipeline. It deliberately does not discuss history, nor better options, as the cost of modifying the existing system is insurmountably high.

The main purpose of this document is to allow other parties to determine whether a newly specced BAO-like structure can represent this existing construction.

The document is complete, but written horribly in a hurry. Shame upon the author's house.

	#!/usr/bin/env perl

	use warnings;
	use strict;

	use File::Find qw/find/;
	use Data::Dumper::Concise;

	my $author_stats;

	# As amply documented at:
	#
	# https://github.com/dbsrgits/dbix-class/blob/dc7d89911/lib/DBIx/Class/MethodAttributes.pm#L242-L298
	# https://github.com/dbsrgits/dbix-class/blob/dc7d89911/lib/DBIx/Class/ResultSourceProxy.pm#L137-L143
	# https://github.com/dbsrgits/dbix-class/blob/dc7d89911/lib/DBIx/Class/ResultSourceProxy.pm#L191-L262
	#


	perl -e '
	{

	# In your base Result class ( ...::Schema::BaseResult )

	sub add_relationship {
	my $self = shift;
	my $opts = $_[3] \|\| {};

	# shut off all DBIC/perl-side cascading logic
	$opts->{cascade_delete} = 0;
	$opts->{cascade_update} = 0;
	$opts->{cascade_copy} = 0;

	# During space-race-1 we tuned the dealbots frequency to not overwhelm the miners
	# With space-race-2 underway your miner will be negotiating deals with a number of
	# clients, who lack coordination among themselves. Therefore it is very strongly
	# recommended to tune your miner parameters based on the sealing capacity you already
	# known from space-race-1
	#
	# The numbers below are the bare minimum for a single-box, non-gpu miner, and are
	# very conservative safe numbers. No doubt your miner can do better so you can bring
	# these numbers up. But! Do not dial them too high: if you lose sectors due to an
	# unresponsive miner, your clients will be... sad.

	#!/bin/bash

	set -eu
	set -o pipefail

	# derive API_INFO from implied defaults if necessary
	LOTUS_PATH="${LOTUS_PATH:-$HOME/.lotus}"
	LOTUS_CFG_MADDR="$( cat $LOTUS_PATH/api 2>/dev/null \|\| true )"
	FULLNODE_API_INFO="${FULLNODE_API_INFO:-${LOTUS_CFG_MADDR:-/ip4/127.0.0.1/tcp/1234/http}}"

Peter Rabbitson ribasushi

Note!!!

Filecoin CommD