Jason Buscema

Here is a very simple script I wrote to create snapshots of all attached EBS volumes for a specific instance. I call this script via Cron on a regular basis. The script is available for download at http://pastie.org/3027416 You will need to have the Amazon EC2 API Tools installed on the same machine this script is executed from.

#!/bin/bash
#
# Created by Jason Buscema <jbuscema@lambesis.com>
#
# Create snapshots of the attached EBS volumes for a specific instance. Script
# will look for ALL volumes assigned to an instance and create snapshots of
# them.
# You will need to have the Amazon EC2 command line tools installed in
# order for this script to work. Get them from http://aws.amazon.com/developertools/351

# Usage
# Call this script with an instance name and a readable name. IE:
# ./backup-s3.sh [instance] [readable name]
# ./backup-s3.sh i-123456 MyServer

# Define our exports we'll need for the tools to work. There are better
# ways of doing this.
export EC2_HOME=/root/.ec2
export PATH=$PATH:$EC2_HOME/bin
export EC2_PRIVATE_KEY=$EC2_HOME/pk-myprivatekeynamegoeshere.pem
export EC2_CERT=$EC2_HOME/cert-mycertnamegoeshere.pem
export PATH=~/.ec2/bin:$PATH
export JAVA_HOME=/usr/java/default
export PATH=$PATH:/usr/java/default

# Which instance are working with? Read from stdin
INSTANCE="$1"
DESCR="$2"

if [  $1 -a $2 ]
	then

	# Get attached volumes for specific instance
	TMPFILE="/tmp/`date +%N`"
	ec2-describe-volumes | grep $INSTANCE | awk '{print $2}' > $TMPFILE

	while read line
	do
		echo -e "Creating snapshot for - $DESCR for $INSTANCE from $line (via `hostname`)"
	    ec2-create-snapshot $line -d "$DESCR snap for $INSTANCE from $line (via `hostname`)"
	done < $TMPFILE

	# Remove our tmpfile
	rm -rf $TMPFILE

	exit 1

else
	echo "> ERROR - Need to pass in the instance ID and a name separated by a space. IE: 'backup-s3.sh i-1234567 Name'"
	exit 1
fi

I should find this shockingly disturbing however for some reason I’m not the least bit surprised. That “some reason” is because the people providing the means to jack someone’s cell number is none other than our friends at AT&T.

Think of someone you really hate … now hope they use AT&T so you can steal their mobile number, then lock their account. From this point on, we’ll refer to this person as “the sucker”.

All you need is a business account with AT&T. Sure, to get a business account you will need to give them your federal tax id number and undergo a credit check but that’s not really a big deal. Once you have an AT&T business account, you can take over the sucker’s personal AT&T number without their prior permission or knowledge. The only two pieces of information you will need to know is their mobile number and their name. Real tough to get I know.

How to do the deed:

1. Call AT&T’s business line (Assuming you can stomach calling AT&T in the first place)
2. Tell them you want to transfer a personal account/number to your business account
3. Give them the super secret information (the sucker’s name and mobile number)
4. Answer whatever questions they have about rate plans, etc.
5. That’s it! You’re done.

Now within a short period of time, the sucker’s mobile number will be transferred to your business account. AT&T will not make any attempt to verify this request with the sucker. Once the number is on your business account, you have full control over the sucker.

Sounds impossible, right? I’ve done this now with three different employees of ours over the past week. In my case, I told two of them what I was doing but I didn’t have to verify anything with AT&T. The third employee I simply forgot to tell him I was going to transfer his mobile number to our corporate account. When the transfer was done and I informed him, he was shocked and upset that it went through without his permission. I was as well. So I called AT&T business and they told me that business accounts do NOT require verification from the personal user to transfer the account.

INSANE!

One of my bigger technical initiatives this year at work has been to provide and maintain an offsite backup of a large subset of the company’s data. Naturally, this needs to be done within a specific budget and it needs to work. While I’m still in the middle of testing, I thought I would share some of my experience and insights. Please keep in mind that the offsite backup solution for us is NOT our first means of backup. We are using Atempo Time Navigator internally and backing up our data to disk. Our archives are also being pushed to LTO tape. Offsite backup for us is a second copy of the important data outside of our office.

Reasons for us to have an offsite backup:

• For diaster recovery (DR) purposes. Theft, fire, flood, earthquake, nukes and gremlins.
• Automation – Reduce the reliance on staff for swapping backup tapes in the libraries and ultimately eliminate any human input and error.
• Equipment failure – A backup to the backups should the backups fail!

Problems with offsite backups:

• Large data set – Big files means more time/bandwidth to transfer offsite. We’re on a DS3 so no issues with bandwidth… yet.
• Security – Where are my files stored and who can get access to them?
• Restore time – In the event of a disaster, how long would it take me to get my data?

Right now I’m testing a very simple solution using Haystack Software’s Arq for Mac and Amazon S3. A license for Arq is $29. Arq is running on of my Mac OS X servers which happens to have a few RAID units connected to it. I’ve simply told Arq to backup a few specific share points on an hourly basis. I’m able to setup excludes which is important.

Arq allows me to set a monthly budget for Amazon S3 and it will automatically purge old version when that budget is reached. Currently, I am using Amazon S3 Reduced Redundancy Storage which costs ¢10/GB. My current backup set is around 4TBs which costs approximately $410/mo.

arq_restore is a command-line utility for restoring from backups created by Arq. Which this command, I can restore all my data from Amazon S3 without the need for the Arq application. This ensures that my data is accessible should Arq ever become unsupported and vanish from this planet.

I’m also using the Amazon EC2 API tools, s3sync and Amazon S3 for automated offsite storage of data from our various Linux servers. For example, we have a very important internal MySQL database that needs to be backed up a few times per day and archived. The database is dumped and backed up via cron every 4 hours and a copy is sent over SSL to a secure Amazon S3 bucket where it sits indefinitely. Very simple and hands free.

So far what I really like about using Arq and Amazon S3 is the simplicity, cost and ease of use.

More to come…

We’ve been having issues all day with Facebook, specifically with the “Like” button. At first we just thought something was wrong with our code when a page was taking nearly 2 minutes to load. Not the case. Unlike Facebook Like.

I’ve been a bit of critic pre-launch of the Apple iPad but thought I would reserve my judgement until after I’ve had the chance to really use it. It’s been a full week now and I like it much more than I did the first day but I’m still not entirely sold on this version… yet.

The iPad definitely doesn’t have the same WIFI range as my MacBookPro but so far I found it to be tolerable. This past week I’ve used the iPad to check news and email, follow a recipe using the Epicurious app (very nice) and use the Pandora Radio app to get some tunes in my room. It was nice not to have to bring my MacBookPro into the kitchen while I was cooking… the iPad was perfect for that.  The ABC Player app is very nice although I don’t happen to watch any ABC shows. My three year old son loves the FirstWords: Animals app. It’s a fun device and will be great on our plane trip to Seattle in a few months. It’ll keep the little ones busy! Although the iPad won’t come close to replacing my MacBookPro, I would consider leaving the now bulky laptop at home and taking the iPad along on future trips. The iPad, as of now, is not a device I will pack along with me daily as I do with my laptop. I’ll mostly use it at home for now. We’ll see.

What’s good about it:

• The screen - It’s very clear and sharp. Viewing video and photos are stunning and it’s easy to read emails, articles or books.
• Speed – The UI is very responsive, especially when compared to my iPhone 3G. The apps run very well.
• Mail app – The Apple Mail app on the iPad is very simple but quite elegant, especially in landscape mode. It’s great to have messages on the left and the message content on the right unlike the iPhone mail app.
• Sound Quality – The sound quality is quite remarkable for such a little device. We don’t have a stereo in our room (unless you count my laptop) so we’ve been using Pandora on the iPad to pipe in music. Sounds great.

What’s not-so-good:

• Size – It’s a great size for the content on the screen but it’s a bit awkward to hold it in your hands.
• Camera – There isn’t one.
• Typing – This will take me some time to get used to as it did with the iPhone. You can’t hold the iPad and type at the same time unless you use one finger which I can’t imagine anyone doing. The only way to type is to put the iPad down on on your lap or table or somehow prop it up. Typing is a bit strange because the keyboard is obviously smaller than a regular computer and some of the keys are hidden. It’s doable but not anywhere near the speed and efficiency of a computer.
• I’m going to drop it – Seriously, the case is so slick!
• Kids – My two kids love it. My oldest is three years old. The iPad was $500. Just not a good combo.

Apple xServe RAID

One thing is consistent when working in the advertising world… the need for storage, then for even more storage. Once you’ve got the storage situation under control, double it… then maybe you’ll be good for a few more months.

Alright, so I’m exaggerating a little. The thing is no matter how much storage I put in place I always seem to need more. It’s a sick disease. The project files we work on seem to be getting exponentially larger over a short amount of time. I remember when an InDesign package for a larger presentation might top off around 20-30MBs. Then I remember when it soon eclipsed 100MBs. Now I see InDesign package files anywhere from 500MBs to over 5GBs. Then there’s the verbal requirement to “backup everything” which means 10TB of additional storage really means around 25TBs including backup space.

Apple xServe RAIDs and Promise VTrak in rack

I have five very pretty looking Apple xServe RAIDs in the server room. They first debuted in February of 2003 and were discontinued by Apple in February of 2008. Many of the drives in these xServe RAIDs were original 250GB or 500GB drives which have been spinning between 3-5 years. Can you say “end of life”? I had to decide what path I was going to take to meet our ever-growing storage needs. I could ditch the outdated xServe RAIDs and go with all new Promise Technology RAIDs or I could look at refurbishing the xServe RAIDs and breath new life into them.

The xServe RAID uses PATA drives and not the newer SATA drives. PATA drives are much harder to come by in the larger sizes and unfortunately much more expensive on a per GB level. Originally these raid units were around $12,000 for 7TBs (14x 500GB). Eventually the 750GB drives were available and for around the same price point this gave you 10.5TB (14x 750GB).

xServe RAID drive module

A new Promise VTrak E-Class 16x SATA raid is $14,999 MSRP and provides 16TBs of raw data. The Apple xServe RAID is 2Gbps fibre while the Promise is 4Gbps. The xServe RAID can support hardware RAID 0,1,5 and JBOD while the Promise can support RAID 0,1,5,6 and more. xServe RAID is limited to PATA while the Promise is SATA or SAS. Clearly, the Promise VTrak system offers quite a bit more than the xServe RAID which is probably why Apple decided to stop competing in this market.

So once again I turned to eBay where I’ve had much success in the past for work related purchases. I found a vendor selling brand new Seagate 750GB PATA drives for around $150/each. Over the course of this past year I purchase a total of 75 drives from this vendor at a total cost of $11, 250. This gave me enough drives to fully refurbish all five xServe RAIDs with five drives set aside as spares. That’s a total cost of $2100 per RAID giving me 10.5TB of raw storage per RAID.

Upgrading the drives is dead simple for anyone with average computer hardware experience. Here’s the basics:

1. Remove any currently configured arrays from the RAID controller via the GUI tool
2. Pull out the seven drives associated with the above now defunct array
3. Pop the end cap off of the each tray with your fingers
4. Unscrew the four screws under the tray holding the drive in place
5. Remove the IDE cable and the power cable
6. Insert the new 750GB drive, reconnect cables, tighten the screws and replace the end cap
7. Make sure the firmware on the xServe RAID is at 1.5.1 or the RAID will not recognize any drive over 500GB
8. Insert all the drives back into the RAID and make sure the blue light shows up on each
9. Setup a new array to your specifications
10. Make sure to enable background initialization if you plan to use the array immediately

Compaq DPS-450CB-1 power supply

I also bought a spare xServe RAID chassis which contains two power supplies, two controllers and a mid-plan logic board. In addition to the spare chassis I also have two more spare controllers and two more spare power supplies, all purchased from eBay over the past year. Some Compaq systems use the exact same power supply but have a different latching mechanism on them. If you are replacing a dead xServe RAID power supply or wish to have a spare, you can buy the Compaq DPS-450CB-1 and just swap out the latching mechanism. These power supplies run around $30 on eBay vs the Apple ones for over $100.

Each newly refurbished xServe RAID is configured into two RAID5 arrays each with 6 drives and a hot spare. This yields 3.41TB of active storage per RAID controller for a total of 6.82TB of available RAID5 storage per RAID.

I purchased a new Promise VTrak E-Class 16x SATA raid which has become our main production storage RAID. Four of the Apple xServe RAIDs are dedicated to our backup system which is running Atempo Time Navigator and the fifth houses low traffic shares. With the four xServe RAIDs connected to the backup server I now have 27.28TBs of available storage!

The bottom-line is simply that I needed lots more storage but didn’t have the budget to purchase all new RAID units. The 4Gbps fibre channel speed increase that the Promise offers over the xServe RAID would never be utilized in the backup environment where the storage was most needed. The biggest downside to going this direction is that the xServe RAIDs are no longer covered under warranty where-as a new Promise RAID would have full coverage. As long as eBay is around, I’m not concerned in the least bit not having my equipment under warranty.

Refurbishing older equipment gives you these benefits:

• Higher ROI
• Lower TCO
• Keeps the boss and finance folks happy
• Mad eBay skills
• It’s green and helps keep equipment out of landfills

Cheap Apple xServe RAIDs can still be found on eBay as many companies abandon old equipment. Do some digging and see how much it would cost to refurbish one of these RAIDs and give it a new life!

All the rumors point to an eminent release of a tablet device from our friends at Apple. Being a fanboy I can’t wait to see what they come up with but I’m struggling to find a place for it in today’s market. Don’t get me wrong, I’ll be all geeked up when Steve Jobs takes the stage in his blue jeans and black shirt to tell us what we’ve all been missing.

The part that I struggle with is where this device fits in. I would imagine it would be lined up somewhere between the iPhone and the MacBook Air. But what benefit will this new device have that either of the above mentioned devices lack? Is Apple really going to break into a new segment and own it with their original design and software? Usually they do just that. Look at the success of the iPhone for example… There were plenty of smartphones out there before Apple came onto the scene but they all were lacking from a user experience and software standpoint. The iPhone owns the smartphone space and others are struggling to catch up. Saw what you will about the AT&T exclusivity.

Artist rendering of a MacBook Touch (Lori M Grant)

Is the Tablet PC market really a viable market for Apple to own? From a gadget factor, an Apple Tablet would definitely stand out and be something some folks, me, will rush out and purchase. From a size factor it would be need to be much like the MacBook Air and much larger than the iPhone. From a functionality standpoint it would need to include WIFI but also mobile service as well. There is no way from a form factor the Apple Tablet would replace the iPhone and from a power and functionality standpoint, it would most likely be lacking when compared to the MacBook Air. So why bother for Apple to roll it out? Can they really come out with a product and own the weak tablet pc market space?

One big mobile device…

The Apple tablet needs to have mobile capabilities outside of just WIFI. Does this mean another exclusive agreement with AT&T or does Apple looks to Verizon to service this device? As an existing iPhone user I would be ok with a minimal monthly charge to my account, say $10/mo, which would allow me to use either the Apple tablet or my iPhone under my existing AT&T account. The current AT&T 3G won’t cut it for me either on this device. I need more speed and more reliability then currently offered for the iPhone. The biggest downside with the iPhone is unfortunately the AT&T service. I don’t see Apple committing another device to AT&T anytime soon.

Potentially great markets for an Apple Tablet…

Gaming: An Apple tablet device has the chance to be a serious gaming platform. With the support of the already successful iTunes App store, an Apple tablet could become a very serious threat to portable gaming devices. The increased size over the iPhone and iPod Touch would be a welcome improvement for gamers, however go to big and the portability factor is gone.

Entertainment: An Apple tablet would be a great video playback device offering much more real estate than the iPhone and iPod Touch. Again with the iTunes App Store users could purchase and rent movies and shows to watch on the device. This feature on it’s own though isn’t enough to support to device, especially at a rumored $600+ price point. If this functionality was only 10% of what the device really did then the portable DVD player market could see a big dip. With airlines like Virgin America now offering free WIFI, a WIFI enabled tablet, with killer battery life, would be something I would purchase for my travels alone. Flying is boring and I need lots of distractions.

Design/Arts: Assuming the larger screen size and touch capability, this device could be a great tool for artists and designers. The Wacom Cintiq product line offers screens artists can draw on however they are tied to an existing workstation and are not portable. The Apple tablet would most likely need to offer a stylus for the artists in addition to using one’s fingers for movement. Apple has been very against the use of a stylus so I don’t think we’ll see this anytime soon. I’m not convinced designers will want to simply finger paint with this device.

At the rate of sounding like Steve Ballmer, I don’t see many executives latching on to the rumored lack of a keyboard. An onscreen keyboard is nice, aka the iPhone, but most execs seem to still like the tactile feedback of a keyboard. Perhaps Apple takes a page from RIM’s Blackberry SurePress and offers an onscreen keyboard with tactile feedback. The MacBook Air has targeted executives, at least in my realm of advertising and design, and has done pretty well in that segment for Apple.

If the goal is simply to make something cool that some people want, Apple will easily hit the mark with a tablet device. However, I’m not sure an Apple Tablet will give user’s something they really need, or can’t live without, with the existing product line already covering the spectrum quite well. Other than the entertainment realm and pure coolness factor, I don’t see a great space for an Apple tablet device currently. I hope Apple will prove me wrong!

As a geek, I want one now! Do I need it? I’ll let Apple decide that for me. Perhaps the tablet is nothing more than a leapfrog to something else only Jobs and crew have in mind for us mere mortals.