How to get Mac High Sierra Time Machine backups working on an Ubuntu 18.04 server

There are lot's of howtos out there how to get Linux servers to support TimeMachine networked backups for Mac clients. But lately they all stopped working due to latest Mac High Sierra updates. Apple moves away from Netatalk protocol and recommends using SMB. Unfortunately the latest Samba you will get on Ubuntu 18.04 is 4.7, and this does not support so called fruit extensions needed by Mac to have Time Machine writing it's data to the destinations. We need to get Samba 4.8 that was released not so long ago.

So let's start. I'm assuming you do not have any other samba installed. The backup folder will pop up as "backup on <your linux server name>" and you will be using user name backups with the password you'll set on the way. I strongly recommend to read the howto I've posted at the end. It allows for a multi user setup.

Install samba:

sudo apt-get-repository ppa:linux-schools/samba-latest
sudo apt update
sudo apt install samba

This installs samba 4.8 that is a hard requirement to get SMB based TimeMachine backups working.

Let's get the old samba config out of the way

mv /etc/samba/smb.conf /etc/samba/smb.confORG

Now create and edit /etc/samba/smb.conf

[global]
server role = standalone server
passdb backend = tdbsam
obey pam restrictions = yes
security = user
printcap name = /dev/null
load printers = no
socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=524288 SO_SNDBUF=524288
server string = Samba Server %v
dns proxy = no
wide links = yes
follow symlinks = yes
unix extensions = no
acl allow execute always = yes

# Special configuration for Apple's Time Machine
fruit:model = MacPro
fruit:advertise_fullsync = true
fruit:aapl = yes
fruit:time machine = yes

## Definde your shares here
[backup]
path = /space/backups
valid users = backups
writable = yes
durable handles = yes
kernel oplocks = no
kernel share modes = no
posix locking = no
vfs objects = catia fruit streams_xattr
ea support = yes
browseable = yes
read only = No
inherit acls = yes

Create and edit /etc/avahi/services/timemachine.service

<?xml version="1.0" standalone='no'?>
<!DOCTYPE service-group SYSTEM "avahi-service.dtd">
<service-group>
 <name replace-wildcards="yes">%h</name>
 <service>
   <type>_smb._tcp</type>
   <port>445</port>
 </service>
 <service>
   <type>_device-info._tcp</type>
   <port>0</port>
   <txt-record>model=RackMac</txt-record>
 </service>
 <service>
   <type>_adisk._tcp</type>
   <txt-record>sys=waMa=0,adVF=0x100</txt-record>
   <txt-record>dk0=adVN=backup,adVF=0x82</txt-record>
 </service>
</service-group>

Create and edit /etc/rc.local as it looks like that samba 4.8 does not come with startup scripts.

#!/bin/bash
echo " # starting samba from rc.local"
smbd
exit 0

Add the backup directory, backup user and set permissions on the directory. This also sets the password for SMB/CIFS access:

useradd -m backups
smbpasswd -a backups
mkdir -p /space/backups
chown backups /space/backups
chmod 700 /space/backups

And let's start this (no reboot required):

/etc/init.d/avahi-daemon restart
smbd

And there, the new backup destination should show up in all macs connected to this network. REMEMBER KIDS, DO ENCRYPTED BACKUPS!

ps. how to view you mac backups progress:

log stream --style syslog --predicate 'senderImagePath contains[cd] "TimeMachine"' --info

Credits:

https://www.reddit.com/r/homelab/comments/83vkaz/howto_make_time_machine_backups_on_a_samba/

Sim racing and game controllers - part 4 - Arduino leonardo project - TENSIOMETERS

Check the previous entries about the project: part1, part2, part3 and the video and problems.

So I'm still using the Arduino racing pad since 2 years. I'm still racing in LFS, now in Assetto Corsa with it. I've complained about what's bad about it. And now I've solved another issue that I've had - the crappy joysticks. I've used FSR402 tensiometers for both accelerator and braking. This seems much more natural than using the lever sticks. And precise. Now I can drive powerful cars with no ABS and TC in Assetto Corsa :)

Why tensiometers?

When I've started thinking about that crappy joysticks, I've looked at what the game consoles are doing for the trigger buttons on the pads:

• X360 pad uses a potentiometer with clever levers,
• Playstation uses tensiometers,
• X One uses Hall Effect sensors.

I wanted to buy the replacement parts for the X360 to check the levers, but the tensiometers are just cheaper. The Hall Effect sensors seem to be a very cool idea, but you really need to have an idea how to sort out the mechanics.

How does it look:

Bad. Ugly. But works.

I've placed the tensiometers along another edge of this beautiful lunch box. And now I can use both methods of controlling the speed of the car by just remapping controls. Frankly speaking - I haven't used the levers since I've got the tensiometers running.

Wiring FSR402 to Arduino:

Check this for examples of wiring and code: https://learn.adafruit.com/force-sensitive-resistor-fsr/using-an-fsr. After wiring the tensiometer properly you use it within the code like any other potentiometer.

Video:

Summary:

This just works. I'm using this for gaming weekly or biweekly (you know, 3 kids) for the last 3-4 months.

Using tensiometers for game controlers has another potential application. People making pedals for racing always look for methods to measure travel for acceleration and pressure for braking - to simulate real car hydraulics. All not ridiculously expensive steering wheels on the market use potentiometers for all pedals. And with a very simple mechanism (1 swing, some foam) you can get a pressure measuring tensiometer based pedal - you'll need a bigger tensiometer thou.

Code:

The requirements for having this running are described in the part3 of this "journey".

#include "HID-Project.h"

const int pinLed = LED_BUILTIN;

const int pinButton = 2;

// this is for printing to COM - set to 1 to get printouts

const int printuj = 0 ;

void setup() {

  pinMode(pinLed, OUTPUT);

  pinMode(pinButton, INPUT_PULLUP);

  Gamepad.begin();

  if ( printuj == 1 ) {

    Serial.begin(9600);

  }

}

void loop() {

  // limits for steering

  float steeringmin=0;

  float steeringmax=1023;

  // limits for joystick braking

  float brakemin=0;

  float brakemax=1023;

  // limits for joystick accelerating

  float accelmax=1023;

  // limits for tensometer acceleration and braking

  float brakemaxb=910;

  float accelmaxb=910;

  float accelmin=0;

  float min=-32767;

  float max=32767;

  float diff=max-min;

  float minB=-123;

  float maxB=+123;

  float diffB=maxB-minB;

  // steering

  float steer = (float)analogRead(0)/(steeringmax-steeringmin)*diff + min;

  // acceleration, braking for joysticks and for tensometers

  float accel = ((float)analogRead(2)/(accelmax-accelmin))*diff + min;

  float brake = ((float)analogRead(1)/(brakemax-brakemin))*diff + min;

  float accelb = ((float)analogRead(4)/(accelmaxb-accelmin))*diffB + minB;

  float brakeb = ((float)analogRead(3)/(brakemaxb-brakemin))*diffB + minB;

  // all 4 buttons readouts

  int gearup = digitalRead(2);

  int geardown = digitalRead(3);

  int gearupb = digitalRead(4);

  int geardownb = digitalRead(5);

  // idiotic deadzones for joystick

  if ((accel <= 3) and (accel >= -3)) {

    accel=0;

  }

  if ((brake <= 3) and (brake >= -3)) {

    brake=0;

  }

  // limits for tensometers

  if (accelb > maxB) {

    accelb=maxB;

  }

  if (brakeb > maxB) {

    brakeb=maxB;

  }

  // this is the COM printing

  if ( printuj == 1 ) {

        Serial.print ((int)steer);

        Serial.print (' ');

        Serial.print (analogRead(0));

        Serial.print (' ');

        Serial.print ((int)accel);

        Serial.print (' ');

        Serial.print (analogRead(2));

        Serial.print (' ');

        Serial.print ((int)brake);

        Serial.print (' ');

        Serial.print (analogRead(1));

        Serial.print (' ');

        Serial.print (gearup);

        Serial.print (' ');

        Serial.print (geardown);

        Serial.print (' ');

        Serial.print (gearupb);

        Serial.print (' ');

        Serial.print (geardownb);

                Serial.print (' ');

                Serial.print (analogRead(3));

                Serial.print (' ');

                Serial.print ((int)brakeb);

            Serial.print (' ');

                Serial.print (analogRead(4));

                Serial.print (' ');

                 Serial.print ((int)accelb);

        Serial.println (' ');

  }

 \\ button switching

 if (gearup == 0)

 {

   Gamepad.press(1);

 } else {

   Gamepad.release(1);

 }

 if (geardown == 0)

 {

   Gamepad.press(2);

 } else {

   Gamepad.release(2);

 }

 if (gearupb == 0)

 {

   Gamepad.press(3);

 } else {

   Gamepad.release(3);

 }

 if (geardownb == 0)

 {

   Gamepad.press(4);

 } else {

   Gamepad.release(4);

 }

 \\ output to axises. xA is 16bit, the rest I don't remember and it does not matter.

 Gamepad.xAxis((int)steer);

 Gamepad.yAxis((int)accel);

 Gamepad.rxAxis((int)brake);

 Gamepad.zAxis((int)accelb);

 Gamepad.rzAxis((int)brakeb);

 Gamepad.write();

}

Next step: getting some better chassis for the device.

Sim racing and game controllers - part 3 - Arduino leonardo project

The issue

I've posted these articles on the Arduino racing pad that I've built: part1, part2 and the video and problems. I've complained a bit on the fact that what I have done has only 8bits of information for the steering wheel. That should be enough for steering, but when driving straight in LFS, you could see that the steering wheel "jumps" between positions. 8bits of information is 256 possible positions of the wheel and when you get 270 degrees of rotation on a potentiometer, that is visible - you cannot turn the wheel by a half of a degree.

My third kid started sleeping better, so now I have bits of time here and there to play around with toys. After a beer or two I've said to my self "8bits steering improvement - how hard can it be?".

The ingredients

Back then, one year ago, you needed to copy the custom HID.cpp and USBAPI.h files into the Arduino IDE libraries to get the Joystick definition. Now you don't have to. Arduino 1.6.8 IDE allows for libraries to be added and guy named Nico prepared marvelous library called Arduino HID Project 2.4.3. This library contains Gamepad definition with 32 buttons, 2 dpads, 4 16bit axis and 2 8bit axis.

So I've used:

• Arduino Leonardo with the tupperware racing pad I've build in the previous posts, 
• Arduino IDE 1.6.8,
• Arduino HID Project 2.4.3,
• moderate amount of alcohol.

The solution

The new Arduino HID project makes it very tidy and simple to set up the controler I needed. The code for the Arduino Sketch is given below. When someone will try to copy and play with that, be sure to check which axis has how many bits. X, Y, RX, RY are 16bits, Z and RZ are 8bits.

The Arduino HID needs to be downloaded and copied into the libraries directory of your Arduino IDE installation. I'm sure there are other places you can put it in, but I guess I'm too lazy :)

The results

Now I've got all axis on 16 bits. Arduino Leonardo A/D has 10 bits resolution, so that is an overkill. The USB device is 64 times more precise than what Arduino can return, but it does not matter. I got what I needed on a very simple set of libraries. Now when I turn the wheel knob everything is super smooth. 

Now the only thing keeping me from winning every LFS game is my set of mediocre virtual driving skills.

Next improvement - some other mechanical potentiometers for the throttle and brakes.

The Code

#include "HID-Project.h"

const int pinLed = LED_BUILTIN;
const int pinButton = 2;

void setup() {
  pinMode(pinLed, OUTPUT);
  pinMode(pinButton, INPUT_PULLUP);
  Gamepad.begin();
}

void loop() {
  float steeringmin=0;
  float steeringmax=1023;
  float brakemin=0;
  float brakemax=1023;
  float accelmax=1023;
  float accelmin=0;
  float min=-32767;
  float max=32767;
  float diff=max-min;


  float steer = (float)analogRead(0)/(steeringmax-steeringmin)*diff + min;
  float accel = ((float)analogRead(2)/(accelmax-accelmin))*diff + min;
  float brake = ((float)analogRead(1)/(brakemax-brakemin))*diff + min;
  // on second thought, I just could have multiplied one thing by 64 to get the result
 
  int gearup = digitalRead(2);
  int geardown = digitalRead(3);
 
  if ((accel <= 3) and (accel >= -3)) {
    accel=0;
  }
  if ((brake <= 3) and (brake >= -3)) {
    brake=0;
  }



 
 if (gearup == 0)
 {
   Gamepad.press(1);
 } else {
   Gamepad.release(1);
 }

 if (geardown == 0)
 {
   Gamepad.press(2);
 } else {
   Gamepad.release(2);
 }
 Gamepad.xAxis((int)steer);
 Gamepad.yAxis((int)accel);
 Gamepad.rxAxis((int)brake);
 Gamepad.write();

 // zAxis is 8bit, I wanted to use 16bits on all

}

My first Arduino racing pad - video and conclusions

Hi!

For future reference, check out the video with the Arduino racing pad, that I've built here: http://jakkul.blogspot.com/2015/06/sim-racing-and-game-controllers-part-2_3.html

My first Arduino racing pad in full glory!

There are some problems with the design that need working on:

• The ergonomics are not the best. It's hard to use one hand for acceleration, brakes and gears.
• I think, that 8bits of data fed by Arduino to the PC for steering are not enough.
• You cannot drink beer while driving. Two hands need to be always on the pad. This is not the case when driving with a mouse or a wheel.
• The gaming-pad-like joysticks are crap. About half of the axis travel is analogue, the rest is in full on state. 

I have some ideas for the next design. But I guess that will have to wait until autumn, since my 3rd kid is due to be born within next 2-3 weeks :)

Sim racing and game controllers - part 2 - Arduino leonardo project

I've written here about the need to build a customized pad-like controller for my racing games: http://jakkul.blogspot.com/2015/05/sim-racing-and-game-controllers-part-1.html Now I've managed to find 3 hours yesterday to put the thing together. But lets start from the beginning.

The Plan

The goal is to have the right hand operate the steering wheel potentiometer, the left hand's index and middle finger to operate accelerator and brakes, left thumb to operate gears up and down. After a long debate I came up with this:

The Plan

The ingredients

The thing that looked the most like the case for the device was a tupperware-like food container. It's also easily available, easy to drill and put things on and has a low low price of zero.

I also bought:

• Arduino Leonardo (clone I guess),
• two thumb joysticks,
• four buttons (two used),
• one linear 10k potentiometer,
• some cables.

The stuff

Building

Building was a fun and easy part. I've screwed everything to the container, made a hole for cabling and connected together.

 Connecting it all was easy:

• 5V and ground everywhere,
• digital pins 2 and 3 to the gear switches,
• X axis of the thumb joysticks (both) to analog lines 1 and 2,
• potentiometer to analog line 0.

I haven't connected the rest of the stuff available in the joysticks - that was not needed.

Starting up

I've used the following excellent howto on how to get Arduino Leonardo to behave like a Joystick: http://www.instructables.com/id/Add-USB-Game-Controller-to-Arduino-LeonardoMicro/?ALLSTEPS. This is also very helpful: http://www.imaginaryindustries.com/blog/?p=80&cpage=1. There are two files that you need to get into your Arduino IDE environment (HID.cpp and USBAPI.h) that define the Joystick, that is not defined in Arduino IDE standard distribution.

I've used Arduino IDE 1.6.4 and Windows 7 64bit. I'm a linux addict, but since this is to be used in racing games, I wanted to test it straight away.

The step 2 of the howto mentioned above provides a test sketch (Arduino program) that allows for the arduino board to send signals as a joystick to your PC. The code that I've made basing on that and other examples in the net is presented below. You can switch displaying all values on COM port for double checking if it works. You can also tune your potentiometers and joysticks using max-min ranges. Keep in mind, that Arduino reads potentiometers and joysticks, returning and integer ranging from 0 to 1023, and the Joystick class requires the values to be between -127 and 127 for each axis.

Summary

It just works! I can now play my racing games with a full analogue controller without the hassle of a steering wheel. It's a different kind of fun, but still fun :)

The code

int useserial = 1;

float steeringmin = 0;
float steeringmax = 1023;
float brakemin = 501; //0;
float brakemax = 1023;
float accelmax = 1023;
float accelmin = 520; //0;
// 0 steer 1 brake 2 accel
float min = -127;
float max = 127;
float diff = max - min;

void setup() {
  Joystick.begin();
  if (useserial == 1 )
  {
    Serial.begin(9600);
  }
}

void loop() {

  float steer = (max - min) / (steeringmax - steeringmin) * ((float)analogRead(0) - steeringmin) + min;
  float accel = (max - min) / (accelmax - accelmin) * ((float)analogRead(2) - accelmin) + min;
  float brake = (max - min) / (brakemax - brakemin) * ((float)analogRead(1) - brakemin) + min;
  int gearup = digitalRead(2);
  int geardown = digitalRead(3);

  if (steer > max) {
    steer = max;
  }
  if (steer < min) {
    steer = min;
  }
  if (accel > max) {
    accel = max;
  }
  if (accel < min) {
    accel = min;
  }
  if (brake < min) {
    brake = min;
  }
  if (brake > max) {
    brake = max;
  }

  if (useserial == 1 )
  {

    Serial.print ((int)steer);
    Serial.print (' ');
    Serial.print (analogRead(0));
    Serial.print (' ');
    Serial.print (accel);
    Serial.print (' ');
    Serial.print (analogRead(2));
    Serial.print (' ');
    Serial.print (brake); Serial.print (' ');
    Serial.print (analogRead(1));
    Serial.print (' ');
    Serial.print (gearup);
    Serial.print (' ');
    Serial.println (geardown);
  }
  if (gearup == 0)
  {
    Joystick.pressButton(0);
  } else {
    Joystick.releaseButton(0);
  }

  if (geardown == 0)
  {
    Joystick.pressButton(1);
  } else {
    Joystick.releaseButton(1);
  }
  Joystick.setXAxis((int)steer);
  Joystick.setYAxis((int)accel);
  Joystick.setZAxis((int)brake);
  Joystick.sendState();
}

Sim racing and game controllers - part 1 - Arduino leonardo project

So I've always dreamed about getting a steering wheel for my PC and playing LFS. There are three downfalls of this idea:

• Its large - so hiding it from my better half will be hard.
• Its noisy - force feedback on logitech g25/g27 and force GT wheels use some sort of gear that emits a noise while turning the wheel. So playing after my kids are asleep is a no-no. And I would wake up my wife.
• I've got a wife, who would like me not have a steering wheel for my games. There is a term WAF in audio world - the Wife Acceptance Factor.

I even borrowed a Logitech G25 from a friend and didn't like it for one more reason: it lags. You can see, that the wheel in the game does not turn as fast as you do. It lags more than a mouse.

I like playing LFS with my mouse, but there are two issues:

• You cannot gradually add power or brake, so handling more powerful cars than the GTI is difficult.
• Hardly any other racing game provides mouse steering, and I hate driving with keyboard.

I've seen the HIMOTO simulator for getting to know how to drive RC cars. (link from rchobbystore.co.uk) But it's not available here in Poland.

I want a device that is small, can be placed on my desk, would have analog inputs for steering, gas and braking, would have some buttons. And I can't find any. Console pads with analog sticks don't cut it. The steering precision is faaaar from what you can get with a mouse or a cheap wheel. I can live without force feedback.

So I've googled for one hour and just ordered an Arduino Leonardo clone, two analogue sticks for accelerating and breaking, some switches and an analogue 10k linear potentiometer for steering.

Of course I could get an axis for hand brake and clutch and a hat switch and a .... but let's keep it simple, down to the absolute minimum that is needed to play LFS. If this runs, I'll develop later on. SCRUM MVP FTW! This is my first project of the kind anyway.

I will keep posting about the journey.

How to tell if your Intel system might be affected by Cougar Point SATA Bug

Following up on this http://www.anandtech.com/show/4143/the-source-of-intels-cougar-point-sata-bug and this http://blogs.intel.com/technology/2011/01/chipset_design_flaw/ and this http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/6-and-c200-chipset-specification-update.pdf.

There was a big product recall in the beginning of 2011, when Intel pulled back from the market B2 stepping 6-series chipset based motherboards. The problem was in chips marketed as H67, P67, HM67, HM65, Q67, Q65, B65, Z68, UM67, QS67, QM67 and server incarnations C202, C204, C206. The problem was that the 3 Gbps SATA ports stability degraded over time, while the 6 Gbps SATA ports remained fully functional. The link above from Anandtech discuses the case.

The problem can be observed by the following in your dmesg on Linux boxes:

[19115.610095] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[19115.614138] ata3.00: configured for UDMA/33
[19115.614164] ata3: EH complete
[19115.614422] ata3.00: exception Emask 0x40 SAct 0x0 SErr 0x80800 action 0x6
[19115.614489] ata3.00: irq_stat 0x40000001
[19115.614543] ata3: SError: { HostInt 10B8B }
[19115.614598] ata3.00: failed command: IDENTIFY PACKET DEVICE
[19115.614661] ata3.00: cmd a1/00:01:00:00:00/00:00:00:00:00/00 tag 0 pio 512 in
[19115.614664]          res 51/04:01:00:00:00/00:00:00:00:00/00 Emask 0x41 (internal error)
[19115.614819] ata3.00: status: { DRDY ERR }
[19115.614873] ata3.00: error: { ABRT }
[19115.614931] ata3: hard resetting link

The thing to look at are the lines with SError, so:

oshii:~# dmesg | grep SError | tail -n 10
[10873.423785] ata4: SError: { UnrecovData HostInt 10B8B BadCRC }
[10873.426642] ata3: SError: { UnrecovData HostInt 10B8B LinkSeq }
[10873.751681] ata3: SError: { UnrecovData HostInt 10B8B Handshk }
[10873.772417] ata4: SError: { UnrecovData HostInt 10B8B BadCRC }
[10874.094513] ata3: SError: { UnrecovData HostInt 10B8B Handshk }
[10874.430439] ata3: SError: { UnrecovData HostInt 10B8B Handshk }
[10874.764108] ata3: SError: { UnrecovData HostInt 10B8B Handshk }
[10875.090231] ata3: SError: { UnrecovData HostInt 10B8B Handshk }
[19115.294031] ata3: SError: { UnrecovData Proto HostInt 10B8B LinkSeq TrStaTrns }
[19115.614543] ata3: SError: { HostInt 10B8B }

According to many sites around the web, the above means SATA cable errors or SATA power supply errors. Strings ata3 and ata4 are the SATA ports that are experiencing the problem. The above means that this happened on first two 3 Gbps SATA ports on the board, as the kernel enumerates them starting from ata1. Ata1 and ata2 are 6 Gbps ports.

There are tools to check if your system might be affected for Windows, but there are none for Linux.  Those tools check the revision of SATA controller. On Linux thou, the only thing to do is to run lspci and check the revision. 04 means potential problems, 05 means your system is OK.

Example of an affected system:

oshii:~# lspci | grep SATA
00:1f.2 SATA controller: Intel Corporation Cougar Point 6 port SATA AHCI Controller (rev 04)

Example of a system that is OK:

motoko3:~# lspci | grep SATA
00:1f.2 SATA controller: Intel Corporation Cougar Point 6 port SATA AHCI Controller (rev 05)

If you use 3 Gbps SATA ports and have revision 04 of the SATA AHCI Controller on a Sandy Bridge Intel 6-Series chipset, you might want to claim your MOBO under warranty.