Dan Makovec's site

SSO between Google Apps (G Suite) and AWS Console

Thu, 27 Aug 2020 09:00:00 GMT

So I wanted to play with SAML SSO using my Google Apps (G Suite) service as my IDP and my AWS account as the client. I diligently followed the directions at Google to do so, but kept getting this error from AWS whenever I attempted to sign in:

Your request included an invalid SAML response. To logout, click here

I couldn't understand what the hell was going wrong, as I'd base64-decoded my SAML responses and saw everything I was supposed to have in there, until I finally came across this AWS troubleshooting page.

It finally made sense:

The attribute must contain one or more AttributeValue elements, each containing a comma-separated pair of strings

Aha! Back to the Google console, I edited my custom user attributes for my Amazon role. Originally I'd had the ARN of the role I had mapped in there, like so:

I had to concatenate it onto the ARN of the SAML provider I'd set up for Google, comma-separated like so:

OK, so start a new browser session, log into my Google Apps (GSuite) account, and click on the AWS app, then:

Hallelujah, all working! The top right now shows the role name (in this case, I called it "Google", and my email address). I'm in, baby!

Rapidly switch AWS CLI profiles

Thu, 02 Jul 2020 14:00:00 GMT

I often find myself using the AWS CLI with multiple accounts and IAM users, and find switching between them using the --profile option tedious.

Here's a cute little shell function to make handling profiles easier, by setting the appropriate environment variable for all subsequent AWS CLI calls in the current terminal session.

When used with a profile name, it simply sets the AWS_PROFILE environment variable to the profile or your choice. It also allows you to quickly view and check the current profile's user details, in case your profile name isn't enough information to tell you what you need to know.

The code

function awsuser() {
        if [ "$1" = "" ]; then
                AWS_ACCOUNT_ALIAS=$(aws iam list-account-aliases --query "AccountAliases[0]" --output text)
                USER_DETAILS=$(aws iam get-user --output json)
                AWS_USER=$(echo $USER_DETAILS | jq -r .User.UserName)
                AWS_ACCOUNT_ID=$(echo ${USER_DETAILS} | jq -r .User.Arn | sed -e 's/.*:://g' -e 's/:.*//g')
                echo "(${AWS_PROFILE:-default}): ${AWS_ACCOUNT_ID}:${AWS_ACCOUNT_ALIAS} -> ${AWS_USER}"
        elif [ "-l" = "$1" ]; then
                AWS_SHARED_CREDENTIALS_FILE=${AWS_SHARED_CREDENTIALS_FILE:-${HOME}/.aws/credentials}
                grep "\[" ${AWS_SHARED_CREDENTIALS_FILE}
        else
                export AWS_PROFILE=$1
        fi
}

Using it

So let's say you've got ~/.aws/credentials configured as follows:

[devaccount-admin]
aws_access_key_id = ......
aws_secret_access_key = ......

[devaccount-developer]
aws_access_key_id = ......
aws_secret_access_key = ......

[prodaccount-admin]
aws_access_key_id = ......
aws_secret_access_key = ......

[prodaccount-user]
aws_access_key_id = ......
aws_secret_access_key = ......

You can get a quick list of all your profiles:

$ awsuser -l
[devaccount-admin]
[devaccount-developer]
[prodaccount-admin]
[prodaccount-user]

Switch to a specific user profile:

$ awsuser devaccount-admin

Then run with no args to see the profile, account ID and alias, and user ID you're currently running with:

$ awsuser
(devaccount-admin): 123456789012:dmakovecdevaccount -> admin

$ aws iam get-user

{
    "User": {
        "Path": "/",
        "UserName": "admin",
        "UserId": "......",
        "Arn": "arn:aws:iam::123456789012:user/admin",
        "CreateDate": "2020-06-01T07:18:28+00:00",
        "PasswordLastUsed": "2020-07-03T05:14:48+00:00"
    }
}

Going Serverless with Gatsby, Netlify and Contentful

Fri, 17 Apr 2020 06:22:00 GMT

So this is my first post on my new "serverless" site. The old one did me well for 5 years, but the world has moved on since Wordpress and I wanted something that didn't need patching and didn't have ongoing maintenance costs. So, during a bit of downtime from work I decided it was time to take a look at the site and see if it was worth updating it.

I've been using AWS pretty much since it was just SQS and nothing but a bunch of APIs. I worked with it as various new tools like S3 and EC2 got rolled out, and have witnessed the revolution in virtualisation that it kicked off.

While I've been working with it non-stop ever since, my personal website was running on a lonely little neglected EC2 instance running Wordpress for years. I'd patched the software every now and then, and the auto-updates for the OS were applied whenever possible. Every now and then though, the instance would need to reboot, or for whatever reason Apache or MySQL would crap out.

Frankly I couldn't be buggered figuring out what was going wrong and fixing it, as my day job is troubleshooting IT issues, and I really try to stay away from doing that I'm not getting paid. Plus, the server was costing me money. Only a few $ a month (it was just a t2.micro), but still I'd rather the money be in my bank account than Mr. Bezos's.

With the amount that I update the site, running a full LAMP stack seems like overkill, and I'd heard of static site generation frameworks like GatsbyJS a while ago, so I set about rebuilding the site using this tech. I had a couple of wants:

I want to be able to publish blog articles using WYSIWYG or close to it (so not e.g. editing Markdown on Github)
I want complete control to make any template or structural changes to the site without needing a full tool chain on any computer (so editing on Github is acceptable here)
I want to work in Typescript rather than plain ol yucky Javascript
I want a contact form in case anybody wants to reach me via the site
I want the site to run on HTTPS under my domain name
I don't want to worry about patching and security
I don't want to worry about the site going down
I don't want pay for hosting

So I discovered headless CMSs, which basically are hosted services that allow you to write content using their fancy tools and provide a REST API for your application (e.g. Gatsby) to pull the content down in a structured manner and manipulate it at will.

I found Contentful which has a free level for developers, which is perfect for me. I then discovered this article, which gave me a good starting template for Gatsby and a jumping off point for integrating with Contentful content types. It also introduced me to Netlify, which provides a CI/CD service that I can leverage to build my site without the need to fire up my own container or EC2 instance.

I set up my page structures and imported all of my old blog posts from WordPress into my space on Contentful. Then I changed the template component files into Typescript and cleaned up the types. Next, I pushed my site configuration up to a GitHub repo, set up my Netlify webhooks and let fly.

Netlify takes care of automatically running the Gatsby build process whenever I add content on Contentful or commit code changes to GitHub, and hosting the finished product. It even has built in LetsEncrypt support so with a little DNS configuration, I get the HTTPS support I need.

Contentful does a great job at providing a decent markdown editor and place to store the sources of my articles. Arengu hosts a form that sends me an email on submit, and there's an Arengu Gatsby plugin which displays the form within the context of my site.

Codeception: Tell Chrome to automatically accept desktop notifications

Thu, 01 Sep 2016 03:38:00 GMT

If you're building a site that makes use of the handy Web Notifications API and want to test it in Codeception, you might want to tell your browser to enable notifications for the site. I had to fiddle a bit before getting this right, so hopefully it'll help somebody. In your xxx-acceptance.suite.yml, here are a couple of flags:

class_name: AcceptanceTester
modules:
  enabled:
    - Asserts
    - WebDriver:
      url: '/'
      browser: chrome
      restart: true
      capabilities:
      # Accept any JS alert boxes by default
        unexpectedAlertBehaviour: 'accept'
        chromeOptions:
          args:
            - 'enable-strict-powerful-feature-restrictions'
            - 'window-size=1440,900'
            - 'window-position=0,0'
          prefs:
            # Automatically accept desktop notifications
            'profile.managed_default_content_settings.notifications': 1

LetsEncrypt on Amazon Linux

Mon, 29 Aug 2016 12:43:00 GMT

This one's pretty simple.

I started with Ivo Petkov's excellent notes and O-mkar's question and self-answer to get LetsEncrypt up on my EC2 instance, then added a cron job.

TL;DR:

sudo bash
yum install python27-devel git
git clone https://github.com/letsencrypt/letsencrypt /opt/letsencrypt
unset PYTHON_INSTALL_LAYOUT
/opt/letsencrypt/letsencrypt-auto --debug
echo "rsa-key-size = 4096" &gt;&gt; /etc/letsencrypt/config.ini
echo "email = email@example.com" &gt;&gt; /etc/letsencrypt/config.ini
unset PYTHON_INSTALL_LAYOUT
/opt/letsencrypt/letsencrypt-auto certonly --webroot -w /var/www/yourdomainroot -d yourdomain.com -d www.yourdomain.com --config /etc/letsencrypt/config.ini --agree-tos
yum install mod24_ssl

Add the following to /etc/httpd/conf.d/vhost.conf:

<VirtualHost *:443>
ServerName yourdomain.com
DocumentRoot "/var/www/yourdomainroot"
<Directory "/var/www/yourdomainroot">
        AllowOverride All
</Directory>SSLEngine on
SSLCertificateFile /etc/letsencrypt/live/yourdomain.com/cert.pem
SSLCertificateKeyFile /etc/letsencrypt/live/yourdomain.com/privkey.pem
SSLCertificateChainFile /etc/letsencrypt/live/yourdomain.com/chain.pem
SSLProtocol All -SSLv2 -SSLv3
SSLHonorCipherOrder on
SSLCipherSuite "EECDH+ECDSA+AESGCM EECDH+aRSA+AESGCM EECDH+ECDSA+SHA384 EECDH+ECDSA+SHA256 EECDH+aRSA+SHA384 EECDH+aRSA+SHA256 EECDH+aRSA+RC4 EECDH EDH+aRSA !RC4 !aNULL !eNULL !LOW !3DES !MD5 !EXP !PSK !SRP !DSS"
</VirtualHost>

And finally, a renewal cron job:

echo > /opt/letsencrypt/autorenew <<EOF

#!/bin/bash
unset PYTHON_INSTALL_LAYOUT
/opt/letsencrypt/letsencrypt-auto renew --config /etc/letsencrypt/config.ini --agree-tos &amp;&amp; apachectl graceful

EOF
chmod a+x /opt/letsencrypt/autorenew

Then run crontab -e and add the following entry:

0 0 * * * /opt/letsencrypt/autorenew

For bonus marks, since you've probably got HTTP vhost for port 80 something like:

<VirtualHost *:80>
    DocumentRoot "/var/www/yourdomainroot"
    ServerName yourdomain.com
    ServerAlias yourdomain.com

    <Directory "/var/www/yourdomainroot">
    AllowOverride All
    </Directory>
    # Other directives here
</VirtualHost>

Simply add the following into your .htaccess to redirect everybody hitting your formerly insecure site to https:

RewriteEngine On
RewriteCond %{HTTPS} off
RewriteRule .* https://%{HTTP_HOST}%{REQUEST_URI} [L,R=301]

Stop OSX Photos.app opening every time you plug your (Android or IOS) device into your Mac

Fri, 18 Mar 2016 00:44:00 GMT

This has been driving me mad for ages, every time I plugged a test device in to debug a Cordova app. I eventually found this solution on SE for El Capitan:

defaults -currentHost write com.apple.ImageCapture disableHotPlug -bool YES

Problem solved!

How to move Google Authenticator to a New Unrooted Android Phone

Thu, 12 Jun 2014 02:19:00 GMT

I just picked up a new Galaxy S4, since they're going cheap now that the S5 is out.

I've had my trusty S2 for a number of years, and it's been hacked around, rooted, had Cyanogenmod on it and everything else you do to a phone that's out of warranty.

But now it's time to move all of my data to the new phone, I see the problem of moving Google Authenticator between devices still hasn't been made easy.

For clarity, if you're just using Authenticator for Google itself, and you only want one device at a time to use it, then it is indeed pretty easy, and Google makes the process fairly straightforward. I however use Authenticator for 2FA on a number of services, and I have it installed on both my phone and tablet, so my use case is different.

Note that if you can root your new device, it's far easier just to install Titanium Backup on your old and new device and move your Authenticator settings between the two. I however can't root my S4 yet because I don't want to void its warranty by tripping Knox.

So here's what you need on your old phone. I'll assume you know what the following terms mean. Google or xda-developers will help you out if you don't:

The phone must be rooted
Developer options enabled
USB debugging enabled
Android Debug Bridge (adb) installed on your PC/Mac
A USB connection between the computer and phone, with adb shell able to connect

On your computer, you'll need sqlite3 installed. Mac users can install it using Homebrew (recipe: sqlite). Everybody else, you can figure it out.

Now, grab your Authenticator database off of the old phone:

dan@Dans-MacBook-Pro ~ $ adb pull /data/data/com.google.android.apps.authenticator2/databases/databases

Note that if the above step fails with a permission denied error, it's because your phone is locked down (kinda like my new unrooted S4), so you're outta luck.

Now open the database locally on your computer:

dan@Dans-MacBook-Pro ~ $ sqlite3 databases
SQLite version 3.7.11 2012-03-20 11:35:50
Enter ".help" for instructions
Enter SQL statements terminated with a ";"
sqlite&gt; .tables
accounts          android_metadata
sqlite&gt; select * from accounts;
1|Google|aaaabbbbccccdddd|0|0|0||
2|Dropbox|aaaabbbbccccdddd|0|0|0||
3|Facebook|aaaabbbbccccdddd|0|0|0|Facebook|Facebook

You'll see the keys for each of your authenticator accounts shown in the list (shown above as aaaabbbbccccdddd).

Now, grab your new phone, open Authenticator, and one at a time add each account, selecting the "Enter provided key" option.

Once you're done with each account, compare the time code generated on the new device with your old one. If they don't match, you made a typo.

Now of course, since in the past few years Google, Samsung et al have made it progressively harder to get to this information, it might be an idea to back up these codes somewhere safe so that next time you upgrade you've got something to refer to!

I hope that helps someone.

Working with Github pull requests on other people's repos

Mon, 04 Feb 2013 14:00:00 GMT

I'm using Composer for a ZF2 project. I just upgraded to Zend Framework 2.1 from 2.0.6, and found that Ben Youngblood's bjyoungblood/BjyProfiler stopped working. A quick query on Freenode's #zftalk channel (thanks Diemuzi) showed me that a fix and subsequent pull request had been made, but the project's maintainer hadn't pulled it in yet. So how can I make use of this fix?

First step: fork the project on Github (dmakovec/BjyProfiler)
Next, clone it down to my computer. I used the OSX GitHub GUI to do this, but "git clone" would do the trick
Locate the pull request (Diemuzi showed it to me, but you'll find it in the github page if you're not so lucky)

Go into your clone and perform the pull:

cd ~/dev/BjyProfiler && git pull https://github.com/internalsystemerror/BjyProfiler.git hotfix/issue-21

Push it back up - git push origin master
Alter your project's composer.json file to add your project repo in.

...
    "homepage": "http://www.example.com/",
# BEGIN INSERT #
    "repositories": [
        { 
            "type": "vcs", 
            "url": "https://github.com/dmakovec/BjyProfiler" 
        } 
    ], 
# END INSERT #
    "require": {
...

Blow away the project's vendor folder (rm -rf vendor/bjyoungblood) - if you don't do this, composer may not update your repo and you'll be stuck scratching your head as to WTF went wrong for an hour.

What you're doing is telling composer to check the named repositories for any of the packages named in the "require" section before going off to packagist and doing the same. If it finds a match in the repo (in this case, the "BjyProfiler" repo will contain a clone of the "composer.json" file in the top level folder, advising composer that the repo is indeed "bjyoungblood/bjy-profiler"), then it will pull the package from that repo instead of Packagist. Then you just keep an eye on the project so that when the change is merged in, you can go back to the upstream project repo.

Upgrading NetGear Stora NAS drives Without Copying Data

Wed, 23 Jan 2013 02:55:00 GMT

A few years ago, I purchased a NetGear Stora NAS for my home. It's a great device, proving useful for general file storage as well as Time Machine backups and iTunes sharing. Thanks to the guys at OpenStora, it has proven to be an easy machine to customize with additional functionality. DLNA for media playback, a Transmission client for getting CentOS updates via BitTorrent, and even a CrashPlan node for backing up over the net.

I originally bought the machine 3 years ago with a single 1TB drive. Once they became cheap enough I threw a second 1TB drive in for RAID1 security. But over the past few months it started filling up, so I decided to invest in a couple of 2TB drives to get some more life out of it.

The question I had was: how can I upgrade these drives without having to spend forever copying all the files off of it and back on over a network (given I didn't have any 1TB USB drives)?

I eventually found a thread here which got me started. Never having used most of the tools discussed before, I did a little reading around and did the following. The whole process took about a day. I found it so handy, and know a few friends with Storas who might benefit from knowing the process I followed. If it helps you, let me know! The usual disclaimers apply of course :)

Here's the procedure:

Shut down Stora

Swap out right-hand side old drive for a new one. Label the drive in case you need to put it back in.
Boot up
Log into web console and go to Preferences -> Disk Management

The second drive is marked as unconfigured. Click relevant button to add drive to RAID1 array

Wait for rebuild to complete. Status can be seen either on web console or via SSH, cat /proc/mdstat (it will take at least 2 hours)
Once the rebuild is complete, repeat the process for the left-hand drive, i.e:
Shut down Stora
Swap out left old drive for a new one and boot up
Log into web console and go to Preferences - > Disk Management

This time the first drive will marked as unconfigured. Click relevant button to add the drive to RAID1 array

Wait for rebuild to complete. Status can be seen either on web console or via SSH, cat /proc/mdstat (it will take at least 2 hours)

At this point, both 2TB drives will have taken over from the old 1TB drives in serving RAID, but they will show only 1TB capacity.

ssh into Stora and get root:

-bash-3.2$ sudo bash

We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these three things:

#1) Respect the privacy of others.
#2) Think before you type.
#3) With great power comes great responsibility.

Password:
audit_log_user_command(): Connection refused
bash-3.2#

Take a look at the disk space on the RAID. Check the row for /home:

bash-3.2# df -m
Filesystem 1M-blocks Used Available Use% Mounted on
rootfs 212 158 54 75% /
ubi0:rootfs 212 158 54 75% /
none 62 1 62 1% /dev
nodev 62 1 62 1% /var/log
nodev 62 1 62 1% /mnt/tmpfs
nodev 62 0 62 0% /var/lib/php/session
nodev 953799 631836 321963 67% /tmp
nodev 62 1 62 1% /var/run
nodev 62 1 62 1% /var/cache
nodev 62 1 62 1% /var/lib/axentra_sync
nodev 62 1 62 1% /var/lib/oe-admin/minions
nodev 62 1 62 1% /var/lib/oe-admin/actions
nodev 62 1 62 1% /var/lib/oe-update-checker
nodev 62 1 62 1% /etc/blkid
nodev 62 1 62 1% /var/lib/dbus
nodev 62 1 62 1% /var/lib/dhclient
nodev 62 1 62 1% /var/lock
nodev 62 1 62 1% /var/spool
nodev 62 1 62 1% /etc/dhclient-eth0.conf
nodev 62 1 62 1% /etc/printcap
nodev 62 1 62 1% /etc/resolv.conf
/dev/md0 953799 631836 321963 67% /home
/dev/md0 953799 631836 321963 67% /tmp
/dev/md0 953799 631836 321963 67% /var/cache/mt-daapd

In this case, there is a 1TB RAID array that's 67% full.

Now get an idea as to the structure of Stora's RAID configuration. In the example below, it's running RAID1, with the two drives.

sda1 is the single partition on drive sda - the drive on the left of the machine, and sdb1 is likewise for the one on the right.

bash-3.2# /sbin/mdadm -D /dev/md0 # sda1 is LHS, sdb1 is RHS
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 10:34:33 2013
State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.178622

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 8 17 1 active sync /dev/sdb1

The next thing to do is tell the RAID manager that the right-hand side drive is faulty, so that it stops accessing it

bash-3.2# /sbin/mdadm --fail /dev/md0 /dev/sdb1
mdadm: set /dev/sdb1 faulty in /dev/md0

The RHS drive now displays as faulty

bash-3.2# /sbin/mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 10:36:12 2013
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 1
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.178628

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 0 0 1 removed

2 8 17 - faulty spare /dev/sdb1

Now tell the system that the drive has been removed from the array (don't actually physically remove the drive - this is all done via software):

bash-3.2# /sbin/mdadm --remove /dev/md0 /dev/sdb1
mdadm: hot removed /dev/sdb1

The disk is now marked as removed:

bash-3.2# /sbin/mdadm -D /dev/md0 # The RHS drive no longer shows as part of the array
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 1
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 10:36:30 2013
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 0
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.178636

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 0 0 1 removed

Now it's time to run fdisk on sdb). This will remove the old 1TB partition and create a new 2TB partition. This doesn't delete any data, so provided the new partition has the same start point, the data will still be accessible when the partition table is written to reflecting the new partition.

Run the following commands:

p: Show the existing partition structure (here it shows a single 1TB partition of type "fd" (Linux raid) starting at cylinder 1
d: Delete partition (as there's only 1 it doesn't prompt you for a number
p: Show that the partition table is empty
n: Create new partition
Primary partition
Partition #1
Default cylinder first and last values to use entire disk. Use hex code "fd" to set partition as Linux RAID type
p: Show that the partition is created as desired
w: Write the new partition table to disk

bash-3.2# /sbin/fdisk /dev/sdb
The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sdb1 1 121577 976562500 fd Linux raid autodetect

Command (m for help): d
Selected partition 1

Command (m for help): p

Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-243201, default 1):
Using default value 1
Last cylinder or +size or +sizeM or +sizeK (1-243201, default 243201):
Using default value 243201

Command (m for help): p

Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sdb1 1 243201 1953512001 83 Linux

Command (m for help): t
Selected partition 1
Hex code (type L to list codes): fd
Changed system type of partition 1 to fd (Linux raid autodetect)

Command (m for help): p

Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sdb1 1 243201 1953512001 fd Linux raid autodetect

Command (m for help): w
The partition table has been altered!

Calling ioctl() to re-read partition table.

WARNING: If you have created or modified any DOS 6.x
partitions, please see the fdisk manual page for additional
information.
Syncing disks.

Now, run fdisk again and use "p" to verify that the partition table was indeed written out correctly.

bash-3.2# /sbin/fdisk /dev/sdb

The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sdb1 1 243201 1953512001 fd Linux raid autodetect

Command (m for help): q

Add the drive back into the array

bash-3.2# /sbin/mdadm --add /dev/md0 /dev/sdb1
mdadm: added /dev/sdb1

Verify that the drive has started rebuilding again

bash-3.2# /sbin/mdadm -D /dev/md0 
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 10:48:02 2013
State : clean, degraded, recovering
Active Devices : 1
Working Devices : 2
Failed Devices : 0
Spare Devices : 1

Rebuild Status : 0% complete

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.178734

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
2 8 17 1 spare rebuilding /dev/sdb1
bash-3.2# cat /proc/mdstat # Get your estimate as to when the rebuild will be finished
Personalities : [linear] [raid0] [raid1] [raid6] [raid5] [raid4]
md0 : active raid1 sdb1[2] sda1[0]
976562432 blocks [2/1] [U_]
[&gt;....................] recovery = 2.2% (21843520/976562432) finish=128.4min speed=123842K/sec

This will take a while (over 2 hours in the example shown above). You can run the above command or simply cat /proc/mdstat to check on progress. When it looks finished, verify that both disks are active again:

bash-3.2# /sbin/mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 13:33:49 2013
State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.181086

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 8 17 1 active sync /dev/sdb1

Now it's time to set up the left hand drive (sda). We just repeat the same procedure as above with sdb:

bash-3.2# /sbin/mdadm --fail /dev/md0 /dev/sda1
mdadm: set /dev/sda1 faulty in /dev/md0
bash-3.2# /sbin/mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 13:35:09 2013
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 1
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.181092

Number Major Minor RaidDevice State
0 0 0 0 removed
1 8 17 1 active sync /dev/sdb1

2 8 1 - faulty spare /dev/sda1
bash-3.2# /sbin/mdadm --remove /dev/md0 /dev/sda1
mdadm: hot removed /dev/sda1
bash-3.2# /sbin/mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 1
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 13:35:33 2013
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 0
Spare Devices : 0

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.181102

Number Major Minor RaidDevice State
0 0 0 0 removed
1 8 17 1 active sync /dev/sdb1
bash-3.2# /sbin/fdisk /dev/sda

The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 1 243153 1953125000 fd Linux raid autodetect

Command (m for help): d
Selected partition 1

Command (m for help): p

Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-243201, default 1):
Using default value 1
Last cylinder or +size or +sizeM or +sizeK (1-243201, default 243201):
Using default value 243201

Command (m for help): t
Selected partition 1
Hex code (type L to list codes): fd
Changed system type of partition 1 to fd (Linux raid autodetect)

Command (m for help): p

Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 1 243201 1953512001 fd Linux raid autodetect

Command (m for help): w
The partition table has been altered!

Calling ioctl() to re-read partition table.
Syncing disks.
bash-3.2# /sbin/fdisk /dev/sda

The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 1 243201 1953512001 fd Linux raid autodetect

Command (m for help): q

bash-3.2# /sbin/mdadm --add /dev/md0 /dev/sda1
mdadm: added /dev/sda1
bash-3.2# /sbin/mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Sun Aug 15 13:26:52 2010
Raid Level : raid1
Array Size : 976562432 (931.32 GiB 1000.00 GB)
Used Dev Size : 976562432 (931.32 GiB 1000.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Tue Jan 22 13:42:22 2013
State : clean, degraded, recovering
Active Devices : 1
Working Devices : 2
Failed Devices : 0
Spare Devices : 1

Rebuild Status : 0% complete

UUID : f1330325:7aaf8fc4:c524c7d5:22e065bd
Events : 0.181250

Number Major Minor RaidDevice State
2 8 1 0 spare rebuilding /dev/sda1
1 8 17 1 active sync /dev/sdb1
bash-3.2# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid6] [raid5] [raid4]
md0 : active raid1 sda1[2] sdb1[1]
976562432 blocks [2/1] [_U]
[&gt;....................] recovery = 0.6% (6342592/976562432) finish=115.3min speed=140204K/sec

unused devices: &lt;none&gt;

So now it's a few hours later and the RAID is up and running again with the two disks now re-partitioned. The RAID itself however still thinks it's operating with the original disk size. So tell it to grow to fill the entire available space. Again, this will take a few hours.

bash-3.2# /sbin/mdadm --grow /dev/md0 -z max

Growth progress can be checked with cat /proc/mdstat, or alternatively, use the --wait option to pause the command prompt from executing the next command until RAID growth is complete:

bash-3.2# /sbin/mdadm --wait /dev/md0

Once it is complete, the filesystem on the virtual disk needs to be grown to fill it. This is a relatively quick operation.

bash-3.2# /usr/sbin/xfs_growfs -D max /dev/md0
meta-data=/dev/md0 isize=256 agcount=32, agsize=7629394 blks
= sectsz=512 attr=0
data = bsize=4096 blocks=244140608, imaxpct=25
= sunit=0 swidth=0 blks, unwritten=1
naming =version 2 bsize=4096
log =internal bsize=4096 blocks=32768, version=1
= sectsz=512 sunit=0 blks
realtime =none extsz=65536 blocks=0, rtextents=0
data blocks changed from 244140608 to 488377984

Now take a look at mounted disk space. /home should now show plenty of free space!

bash-3.2# df -m
Filesystem 1M-blocks Used Available Use% Mounted on
rootfs 212 158 54 75% /
ubi0:rootfs 212 158 54 75% /
none 62 1 62 1% /dev
nodev 62 1 62 1% /var/log
nodev 62 1 62 1% /mnt/tmpfs
nodev 62 0 62 0% /var/lib/php/session
nodev 1907599 631836 1275764 34% /tmp
nodev 62 1 62 1% /var/run
nodev 62 1 62 1% /var/cache
nodev 62 1 62 1% /var/lib/axentra_sync
nodev 62 1 62 1% /var/lib/oe-admin/minions
nodev 62 1 62 1% /var/lib/oe-admin/actions
nodev 62 1 62 1% /var/lib/oe-update-checker
nodev 62 1 62 1% /etc/blkid
nodev 62 1 62 1% /var/lib/dbus
nodev 62 1 62 1% /var/lib/dhclient
nodev 62 1 62 1% /var/lock
nodev 62 1 62 1% /var/spool
nodev 62 1 62 1% /etc/dhclient-eth0.conf
nodev 62 1 62 1% /etc/printcap
nodev 62 1 62 1% /etc/resolv.conf
/dev/md0 1907599 631836 1275764 34% /home
/dev/md0 1907599 631836 1275764 34% /tmp
/dev/md0 1907599 631836 1275764 34% /var/cache/mt-daapd

DVD Region Unlocking a Samsung HT-E5550W Home Theatre

Sun, 20 Jan 2013 07:09:00 GMT

Ensure that there is no disk in the tray.
Switch the player off and then back on. This is necessary to clean-boot the firmware.
Wait until the main menu page appears and the player stops doing anything.
Press the eject button to open the tray. (The button is between the "Power" and "TV Power" buttons at the top of the controller.)
Press the eject button to close the tray.
Wait while the player searches for a disk.
As soon as the "No Disk" message appears at the top right corner do the following.
Press the REPEAT button – just above the TV channel selector at the bottom right of the controller.
Enter 76884 on the number pad. This is The Code If You Have A Region 4 Player (Other Region Codes Given Below)
The region code "4" should appear at the top right corner of the screen.(this just flashed on then off, so keep an eye on the screen)
If you see "4" over the top right corner, quickly enter "9" to make the player region free.
Power down and power up again.
Play a foreign DVD to verify the change.

Other Region Codes To Be Used In Step 9 If You Dont Have a Region 4 Player

1 – 2 9 3 3 4

2 – 5 7 5 3 8

3 – 5 6 7 3 2

4 – 7 6 8 8 4

5 – 5 3 8 1 4

6 – 2 4 4 6 2