User:Plauchu Edwin: Difference between revisions

From Yocto Project
Jump to navigationJump to search
Line 457: Line 457:
The above model points to an entry in catalog.cfg.
The above model points to an entry in catalog.cfg.


== Runnig AFT with image ==
== Runnig AFT with Ostro image ==


By using sudo it is required to type the following
<pre>
sudo ./test.sh
</pre>
This is the content of script mentioned above..
<pre>
<pre>
#!/bin/sh
#!/bin/sh


# This image has to be hosted over tester user's home directory
TEST_IMAGE=ostro-image-swupd-dev-intel-corei7-64-2016-10-19_05-56-10-build-508.dsk
TEST_IMAGE=ostro-image-swupd-dev-intel-corei7-64-2016-10-19_05-56-10-build-508.dsk
export http_proxy='' && export HTTP_PROXY=''
export http_proxy='' && export HTTP_PROXY=''
`which aft` minnowboardmax $TEST_IMAGE
`which aft` minnowboardmax $TEST_IMAGE
</pre>
</pre>

Revision as of 17:54, 26 October 2016

A MinnowBoard MAX working system

The AFT elements and net

The testable devices are put in an isolated network (AKA AFT net).

Line color code
Blue AC cable
Orange USB cable
Green Ethernet cable
Black Not a cable

AFT.png


The high level testing flow

Once AFT was installed and configure. It'll load its configuration options from '/etc/aft/*.cfg' files and reserve a device by using a lockfile so multiple AFT instances can't use the same device at the same time. Then the device is turned off and then on using the Remote power switch that’s defined in the device config files to reboot the device. When we turn the device on we start to send keystrokes right away from the file determined by service_mode_keystrokes in the AFT configuration files. These keystrokes will go to BIOS and choose the support image to boot from. So for example with basic Minnowboard max as DUT we would send F2-key multiple times to make sure we get to BIOS, then use the arrow keys to go to boot menu and then choose the memory stick as the bootable device and then send Enter. After that we try to connect to the DUT with ssh using the ip in the dnsmasq.leases file. We use the 'id' from device config files to know what ip address to use. When we can connect to the DUT with ssh, we make sure that DUT has booted to 'service_mode' by using 'uname -a' command on the device. After that we flash the testable image and inject ssh-keys to it so we can access it easily when we boot it. When the flashing has been successful we reboot the device. The procedure is the same as before, but now utilizing test_mode_keystrokes for booting the testable OS and 'test_mode' to know we have really booted the right OS. After that we invoke the tests on the DUT and on default turn off the device after everything has completed.

Setup harness server

OS Installation

The Linux distribution installed upon harness server during the creation of this document was Ubuntu yakkety.

Important facts upon installation

During Installation process of Ubuntu It'll be requested the creation of a user, please create the user with the name tester. Additionally the computer name will have to be testing-harness.

When installation has finished. The 'sudo group' has to be enabled as passwordless upon file '/etc/sudoers'.

# Allow members of group sudo to execute any command
%sudo ALL=(ALL) NOPASSWD: ALL

It is important to denote that user tester has been set as part of 'sudo group' by the default Ubuntu installation. With user tester is going to be executed the overall of automated testing.

Ram and Disk distribution

The Ram and Disk details this harness server during the creation of this document were:

root@testing-harness:~# free -m
              total        used        free      shared  buff/cache   available
Mem:          15945         251       12904          54        2790       15311
Swap:         16278           0       16278

root@testing-harness:~# fdisk -l /dev/sda
Disk /dev/sda: 931.5 GiB, 1000204886016 bytes, 1953525168 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: gpt
Disk identifier: B15E7DED-0607-4F32-8072-8CEC8041A1A8

Device          Start        End    Sectors   Size Type
/dev/sda1        2048    1050623    1048576   512M EFI System
/dev/sda2     1050624 1920184319 1919133696 915.1G Linux filesystem
/dev/sda3  1920184320 1953523711   33339392  15.9G Linux swap

root@testing-harness:~# df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            7.8G     0  7.8G   0% /dev
tmpfs           1.6G  9.3M  1.6G   1% /run
/dev/sda2       901G   39G  817G   5% /
tmpfs           7.8G  156K  7.8G   1% /dev/shm
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
tmpfs           7.8G     0  7.8G   0% /sys/fs/cgroup
/dev/sda1       511M  3.4M  508M   1% /boot/efi
tmpfs           1.6G   44K  1.6G   1% /run/user/111
tmpfs           1.6G     0  1.6G   0% /run/user/1000


Firewall configuration

Firewall configuration is not provided, it is recommended that the testing harness is only kept in a trusted network.

To check if no firewall are configured over testing-harness, we can expect a similar output when inquiring iptables...

root@testing-harness:~# iptables -L && iptables -t nat -L
Chain INPUT (policy ACCEPT)
target     prot opt source               destination         

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination         

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination         

Chain PREROUTING (policy ACCEPT)
target     prot opt source               destination         

Chain INPUT (policy ACCEPT)
target     prot opt source               destination         

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination         

Chain POSTROUTING (policy ACCEPT)
target     prot opt source               destination         

Setup Security shell

OpenSSH is a freely available version of the Secure Shell (SSH) protocol family of tools for remotely controlling, or transferring files between, computers. That's why it is required its installation upon testing-harness server.

root@testing-harness:~# aptitude install openssh-server 

Setup NFS

AFT applies Network file system to share /home/tester directory where images and other files used in flashing for PC-devices within AFT net can be found.

NFS is installed ?

To check if the NFS server is installed, we can do...

root@testing-harness:~# dpkg -l | grep nfs-kernel-server
How to install NFS ?

To install NFS these are the required packages...

root@testing-harness:~# aptitude install nfs-kernel-server
The following NEW packages will be installed:
  keyutils{a} libnfsidmap2{a} libtirpc1{a} nfs-common{a} nfs-kernel-server 
  rpcbind{a} 
0 packages upgraded, 6 newly installed, 0 to remove and 459 not upgraded.
Need to get 470 kB of archives. After unpacking 1 884 kB will be used.
Do you want to continue? [Y/n/?]
Sharing /home/tester directory

The tester’s home will be set as shared directory

root@testing-harness:~# cd /etc
root@testing-harness:~# echo '/home/tester 192.168.1.0/24 (crossmnt,ro,root_squash,sync,no_subtree_check)' >> exports

Setup DNS cache and DHCP server for AFT net

DNSmasq provides two services, either of which can be used independently.

  1. DNS service
  2. DHCP service

A local DNS cache can speed up internet browsing because the user's applications will not need to access a domain name server when it looks up a domain name the computer has visited before.

How to install dnsmasq ?

To install dnsmasq these are the required packages...

root@testing-harness:~# aptitude  install dnsmasq
The following NEW packages will be installed:
  dns-root-data{a} dnsmasq dnsmasq-base{a} 
0 packages upgraded, 3 newly installed, 0 to remove and 459 not upgraded.
Need to get 0 B/329 kB of archives. After unpacking 886 kB will be used.
Do you want to continue? [Y/n/?

At the sametime we have set up a simple DNS server. To test it, use your favorite DNS lookup tool pointed at localhost

root@testing-harness:~# nslookup yoctoproject.org localhost
Server:		localhost
Address:	127.0.0.1#53

Non-authoritative answer:
Name:	yoctoproject.org
Address: 140.211.169.56
Configure DHCP server

To enable the DHCP server of AFT net, you will need to give it a range of IP addresses to hand out. The AFT net server would hand out 92 address starting at 192.168.1.2 and ending at 192.168.1.99.

root@testing-harness:~# echo 'dhcp-range=192.168.1.2,192.168.1.99,10m' >> /etc/dnsmasq.conf

To tell dnsmasq which ethernet interface it can and cannot listen on.

root@testing-harness:~# echo 'interface=eno1' >> /etc/dnsmasq.conf

During the creation of this document the interface eno1 is the unique one connected to AFT net.

AFT and its relation with dnsmasq.leases

As of dhcp server is up and running. AFT shall utilize dhcp-server's active ip leasing information to determine device IP address of PC-devices over AFT net. By default it is found upon /var/lib/misc/dnsmasq.leases

root@testing-harness:~# cat /var/lib/misc/dnsmasq.leases

1417132679 78:c5:e5:b4:4d:07 192.168.1.99 minox-fbe390b0bb404da *
1417134678 20:54:76:e4:d8:a5 192.168.1.57 minoy-ab6e5d0a524c9c7d *
......
.......
........ and so on...

Above a common dnsmasq.leases file.

Setup USB stick

Write support image on USB stick

A Debian operative system is used as support image for minnowboard. As of the memory stick is connected, it is required to verify how it was detected by OS. It is simple by using dmesg command.

root@testing-harness:~# dmesg
[354629.272050] usb 3-6: Product: ADATA USB Flash Drive
[354629.272051] usb 3-6: Manufacturer: ADATA
[354629.272052] usb 3-6: SerialNumber: 25C1008050080204
[354629.272594] usb-storage 3-6:1.0: USB Mass Storage device detected
[354629.272701] scsi host7: usb-storage 3-6:1.0
[354630.271607] scsi 7:0:0:0: Direct-Access     ADATA    USB Flash Drive  1.00 PQ: 0 ANSI: 6
[354630.271921] sd 7:0:0:0: Attached scsi generic sg1 type 0
[354630.272050] sd 7:0:0:0: [sdb] 30310400 512-byte logical blocks: (15.5 GB/14.5 GiB)
[354630.272189] sd 7:0:0:0: [sdb] Write Protect is off
[354630.272192] sd 7:0:0:0: [sdb] Mode Sense: 23 00 00 00
[354630.272329] sd 7:0:0:0: [sdb] Write cache: disabled, read cache: disabled, doesn't support DPO or FUA
[354630.278200]  sdb: sdb1 sdb2 sdb3
[354630.279114] sd 7:0:0:0: [sdb] Attached SCSI removable disk

From this point onwards we are ready to record the image over USB stick

root@testing-harness:~# wget -qO- $CI_SUPPORT_IMAGES_URL/minnowUSB.image.tar.gz | gunzip | tar -xOvf - | dd of=/dev/sdb
root@testing-harness:~# sync

Now Reboot the device and from BIOS menu select EFI USB Device from Boot Manager option.

Stuff available and steady configurations on support image

We should find these directories upon /mnt. The first one is utilized to mount the network share directory /home/tester of testing-harness

root@Debian-Minnowmax:~# ls /mnt
img_data_nfs  super_target_root   target_root

We should find firmware required for Realtek Ethernet card that is utilizing minnow-board max

root@Debian-Minnowmax:~# dpkg -l firmware-realtek*

The support image can be accessed over SSH without a password (it was achieved by injecting a known public SSH-key to /root/.ssh/authorized_keys). In other hand, config file contains info about the private key to be used when connecting with other hosts. There is also enabled the ssh access to this support image as root.

root@Debian-Minnowmax:~# ls .ssh
authorized_keys   config   id_rsa_testing_harness.pub   id_rsa_testing_harness
root@Debian-Minnowmax:~# tail -n3 /etc/ssh/sshd_config
# ADDITIONS BY SUPPORT IMAGE INSTALLATION SCRIPT:
PermitRootLogin yes
UsePAM no

Additionally we should find the following debian packages installed.

root@Debian-Minnowmax:~# dpkg -l nfs-common nfs-server bmap-tools openssh-server vim python python-pip vim
bash-completion nano initramfs-tools net-tools ntp locales parted attr tree

Required configurations on support image

Logging on support image

As of support image has booted, it is posible to login as root by typing the password rootme

Mounting harness shared directory '/home/tester'

Edit /etc/fstab to mount remote directory tester at every time when booting operative system.

192.168.1.1:/home/tester /mnt/img_data_nfs nfs    rsize=8192,wsize=8192,timeo=14,intr,nolock,auto
Change network device name from eth1 back to eth0

The interface name of a network device increases if the mac address of the physical network card changes. A common case is when booting a computer with an image created over a different one. It is to fix it by editing '/etc/udev/rules.d/70-persistent-net.rules', (Copy the new mac address to the line of your eth0 rule and delete the new rule for eth1).

# PCI device 0x15ad:0x07b0 (r8169)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:13:20:fe:33:72",
 ATTR{type}=="1", KERNEL=="eth*", NAME="eth0"

Install and configure AFT

This section assumes that the testing harness and USB stick have already been setup.

Installation

AFT has to be installed upon testing-harness hosts by using the root user. AFT can be both installed and updated the same way: By cloning this repository or pulling in the latest changes and executing build and install subcommands with the setup.py installer

root@testing-harness:~# git clone https://github.com/01org/AFT
root@testing-harness:~# cd AFT && python setup.py build  
root@testing-harness:~# python setup.py install

Configuration

Edit platform.cfg

By default the file /etc/aft/devices/platform.cfg come with several sections predefined. Each section was created for a platform family.

Minnowboard belongs to PC platform family. The explanation of each field is the file comments.

[PC]
; PC Family are platforms which have a network interface with
; a fixed MAC-address and usually have a some sort of
; BIOS/EFI menu.

; Utilized by AFT to know ip address of any DUTs over AFT net
leases_file_name = /var/lib/misc/dnsmasq.leases

; keyboard sequence that will be used during configuration
; to check and verify that the keyboard emulator is functional
; (!!LATEST AFT COMMIT HAS REMOVED THIS BASIC HEALTH CHECK!!)
config_check_keystrokes = /home/tester/keystrokes/empty

[Edison]

[BeagleBoneBlack]
leases_file_name = /var/lib/misc/dnsmasq.leases
The keystroke files

Only two keystroke files are required.

  1. One to boot with the support image (AKA booting in 'service/support mode').
  2. Another one to boot with the target image (AKA booting in 'test mode')

During the creation of this document we have stored and created the keystroke files within '/home/tester/keystrokes' (By using tester user we have created the directory)

tester@testing-harness:~/keystrokes$ more service_mode.json 
{
   "ktasks": {
     "seq": [
         {
            "_desc": "EFI_Shell",
            "_inst": "<f2>",
            "_time_gap": "1",
            "_times": "5"
         },
         {
            "_desc": "Enter_EFI",
            "_inst": "<enter>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Skip_startup",
            "_inst": "<esc>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Select usb stick as boot device",
            "_inst": "fs1:",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Enter_device",
            "_inst": "<enter>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "EFI_Location",
            "_inst": "efi\\boot\\bootx64.efi",
            "_time_gap": "0",
            "_times": "1"
         },
         {
            "_desc": "Enter_File",
            "_inst": "<enter>",
            "_time_gap": "2",
            "_times": "1"
         },
         {
            "_desc": "Select first grub option",
            "_inst": "<enter>",
            "_time_gap": "10",
            "_times": "1"
         }
     ]
   }
}
tester@testing-harness:~/keystrokes$ more test_mode.json 
{
   "ktasks": {
     "seq": [
         {
            "_desc": "EFI_Shell",
            "_inst": "<f2>",
            "_time_gap": "1",
            "_times": "5"
         },
         {
            "_desc": "Enter_EFI",
            "_inst": "<enter>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Skip_startup",
            "_inst": "<esc>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Select SD CARD as boot device",
            "_inst": "fs0:",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "Enter_device",
            "_inst": "<enter>",
            "_time_gap": "1",
            "_times": "1"
         },
         {
            "_desc": "EFI_Location",
            "_inst": "efi\\boot\\bootx64.efi",
            "_time_gap": "0",
            "_times": "1"
         },
         {
            "_desc": "Enter_File",
            "_inst": "<enter>",
            "_time_gap": "2",
            "_times": "1"
         },
         {
            "_desc": "Select first grub option",
            "_inst": "<enter>",
            "_time_gap": "10",
            "_times": "1"
         }
     ]
   }
}
Edit catalog.cfg

By default the file /etc/aft/devices/catalog.cfg come with several sections predefined. Each section represents a supported device by AFT. It is required to set the above keystroke files over minnow-board max section. It is also required to set the cutter_type to 'netbootercutter' (It is the cutter type which represents the remote power switch utilized over AFT net)

[MinnowboardMAX]
platform = PC
keyboard_emulator = km232keyboard
cutter_type = netbootercutter
test_plan = iot_qatest
target_device = /dev/mmcblk0
root_partition = /dev/mmcblk0p2
service_mode = Debian
test_mode = yocto
service_mode_keystrokes = /home/tester/keystrokes/service_mode.json
test_mode_keystrokes = /home/tester/keystrokes/test_mode.json
serial_bauds = 115200
Edit topology.cfg

Specific details, such as power cutter slot where minnow-board is connected and minnow-board max address should be set in /etc/aft/devices/topology.cfg

[MINNOWBOARD1]
model = MinnowboardMAX
id = 00:13:20:fe:33:72
netbooter_model = np0801dt
channel = 1 

The above model points to an entry in catalog.cfg.

Runnig AFT with Ostro image

By using sudo it is required to type the following

sudo ./test.sh

This is the content of script mentioned above..

#!/bin/sh

# This image has to be hosted over tester user's home directory
TEST_IMAGE=ostro-image-swupd-dev-intel-corei7-64-2016-10-19_05-56-10-build-508.dsk
export http_proxy='' && export HTTP_PROXY=''
`which aft` minnowboardmax $TEST_IMAGE