royling

Install tmux

brew install tmux

Start tmux (https://gitlab.com/gnachman/iterm2/wikis/TmuxIntegration)

Run tmux -CC or tmux -CC attach in iTerm2 and then menu is shown on terminal:

• ESC to detach
• C to run commands
• X to force quit
• L for log

zeeshanlakhani

Testing Stateful Code

So far, all the properties we have written have tested stateless code. Stateless code is made up of pure functions and is inherently easier to test than stateful code with side effects. The chief problem with testing stateful code is that the input to output mapping depends on the current state of the program. Previous operations can cause the same function to return different output given the same input. Therefore, in order to test stateful code, our tests must maintain some state of their own. This state is known as the model state and is updated as part of the testing process.

rbranson

The point of this is to use cheap machines with small/slow storage to coordinate client requests while dedicating the machines with the big and fast storage to doing what they do best. I found that request coordination was contributing to about half the CPU usage on our Cassandra nodes, on average. Solid state storage is quite expensive, nearly doubling the cost of typical hardware. It also means that if people have control over hardware placement within the network, they can place proxy nodes closer to the client without impacting their storage footprint or fault tolerance characteristics.

This is accomplished in Cassandra by passing the -Dcassandra.join_ring=false option when the process is started. These nodes will connect to the seeds, cache the gossip data, load the schema, and begin listening for client requests. Messages like "/x.x.x.x is now UP!" will appear on the other nodes.

There are also some more practical benefits to this. Handling client requests caused us to push the NewSize of the heap up

takeshixx

#!/usr/bin/env python2
"""
Author: takeshix <takeshix@adversec.com>
PoC code for CVE-2014-0160. Original PoC by Jared Stafford (jspenguin@jspenguin.org).

Supportes all versions of TLS and has STARTTLS support for SMTP,POP3,IMAP,FTP and XMPP.
"""

import sys,struct,socket
from argparse import ArgumentParser

bsparrow435

API

HTTP

Entry point for all object operations: https://github.com/basho/riak_kv/blob/1.4.2/src/riak_kv_wm_object.erl

delete_resource/2 takes RequestData(Request header, ex: vclock) and Context(Record containing: Bucket, Key, Client): https://github.com/basho/riak_kv/blob/1.4.2/src/riak_kv_wm_object.erl#L888

kevsmith

-module(bench).

-export([run_suite/3, run/1]).

run_suite(Name, Passes, Messages) ->
    Results = run_suite1(Passes, Messages, []),
    Avg = lists:sum(Results) / Passes,
    StdDev = math:sqrt(lists:sum([math:pow(X - Avg, 2) || X <- Results]) / Passes),
    Summary = [{avg, Avg}, {stddev, StdDev}, {min, lists:min(Results)},
               {max, lists:max(Results)}],

seancribbs

riak_object:bucket/key types should be in different module

riak_client:put increments vclock

riak_index is riak_object only

riak_kv_backup

• Move serialize/deserialize into crdt/riak_object

riak_kv_delete

• vclock used directly
• tombstone logic should be in riak_object
• quora stuff is duplicated from other modules

riak_kv_encoding_migrate is from 0.14 or 1.0, can we remove?

gburd

#!/usr/bin/env bash


# Note: erlang depends on ncurses, openssl at a minimum

usage ()
{
    echo "usage: $0 <release> <type>"
    echo "        release: R14B01|R14B02|R14B03|R14B04|R15B|R15B01|R15B02|R15B03|R16B|R16B01|R16B02"
    echo "        type:    normal, opt, gcov, gprof, debug, valgrind, or lcnt"

hellerbarde

Latency numbers every programmer should know

L1 cache reference ......................... 0.5 ns
Branch mispredict ............................ 5 ns
L2 cache reference ........................... 7 ns
Mutex lock/unlock ........................... 25 ns
Main memory reference ...................... 100 ns             
Compress 1K bytes with Zippy ............. 3,000 ns  =   3 µs
Send 2K bytes over 1 Gbps network ....... 20,000 ns  =  20 µs
SSD random read ........................ 150,000 ns  = 150 µs

Read 1 MB sequentially from memory ..... 250,000 ns = 250 µs

retronym

class RichOrdering[A](oa: Ordering[A]) {
  def thenBy[B: Ordering](f: A => B): Ordering[A] = new Ordering[A] {
    def compare(a1: A, a2: A) = oa.compare(a1, a2) match {
      case 0 => Ordering[B].compare(f(a1), f(a2))
      case x => x
    }
  }
}

implicit def ToRichOrdering[A](oa: Ordering[A]): RichOrdering[A] = new RichOrdering(oa)