I think it has it’s place in between smart phones and laptops, but it still has to rival netbooks. Mileage may vary depending on what you want it to do of course. Personally, I will wait for the HP Slate to come out and get the whole Windows 7 touch experience which I am already accustomed to on my older MotionComputing slate. And I still despise the whole Apple “thinking of your best interest” restrictiveness of the app store and the iphone/ipad platforms. There’s just some things I want to be able to do that a simple app won’t cover, even after jail-breaking.

I had to pick up a couple of suction cups, but the overall tear down process to the iMac was pretty easy. I couldn’t find any instructions on taking apart a 24″ iMac specifically, only a 20″, so I winged it. Just pull the glass off, unscrew and remove the front cover, then unscrew the screen and pull it out of the way. The drive sits right behind the screen and removing that was probably the hardest part. It’s held in place with a plastic flange that you have to kind of push and pull to pop the drive out, then just unplug it.

The next difficult part was putting the SSD in. Given that it is a 2.5″ drive instead of 3.5″, I had to get an adapter to make it fit in place.

Here in lied the rub.. Since it’s a smaller drive, the SATA cable wasn’t long enough to reach the drive, and it wouldn’t budge.

I couldn’t swap it out either because it threaded under and around to the bottom of the motherboard. Fortunately, from what I have seen in the pictures, the new 27″ iMac looks a little more forgiving. So, to solve this, I had to jury rig the drive onto the adapter only using one screw, and given that it’s an SSD, I wasn’t too worried. After that, it was pretty easy to put everything back together.

Before I put the glass on, I wanted to piece it together and do a quick test with the Snow Leopard install disk to make sure the drive actually worked, which it did without a hitch.

This was probably a couple of weeks ago that I did this and so far everything has been working great! Apps don’t launch immediately, but usually within two bounces, and my virtual XP machine runs a lot faster than it did. Next two having two screens, running with an SSD is my next productivity improvement recommendation, I highly recommend it for anyone that can afford it!

However, if you do encounter the above error anyway, you need to restart the VMware network bridge. You can do this either by shutting down VMware completely (turn off your guest operating systems, and quit the VMware Fusion application), or you can run the following commands as an administrator in Terminal, which will stop any bridge currently running (or do nothing if no bridge is running) and then restart it, providing the output as shown:

sudo killall vmnet-bridge
sudo "/Library/Application Support/VMware Fusion/vmnet-bridge" -D vmnet0 ''
Entering event loop...
Examining network configuration...
Turning on bridge with host network interface en1...

Obviously, you may be asked for your password as you perform this procedure. Note that the trailing two apostrophes are single quotes with no space. This is (almost) how the VMware Fusion boot.sh script starts and stops the network bridge. Specifically, you’re telling the vmnet-bridge application to run in Debug mode and to bridge vmnet0 to whatever is the current primary networking interface. In the example output shown above, this is en1, or my AirPort card connected to the computer-to-computer network I created in the previous step.

When I ran the command, I didn’t get any feedback but it was working. Odd. Something for me to report.

To start off, we wanted to do a clean upgrade and leave as much cruft behind as we could. This meant not bothering with importing the old Active Directory and the arduous process it entailed. We simple backed up what we needed, shut down the old SBS2003 server, and fired up the new SBS2008 server. We only had about a dozen accounts to create, only half of those being internal Exchange enabled accounts, so not a big deal. This also made it easier on me since I could go ahead and install the OS (a couple solid hours in itself), and preconfigure the intranet sites we would be using in advance instead of doing it on site like I would have if we did the “proper” upgrade.

Next I had to bring up the new fileserver. I had originally planned on using a hacked version of Home Server but dumped the idea at the last minute because, well, it wasn’t a very sound idea even if it worked. Instead, I opted to go with the tried and true OpenFiler OS running on a 4GB CF card using a CF to IDE adapter. I also dropped in a gigabit card and directly connected it to the second gigabit port on the Dell 2850 server and mounted it via ISCSI with the help of some instructions from their FORUM. Again, this is a four 250GB drive RAID10 array and so far it’s been running great. All of the shared folders and user “My Documents” are redirected to this server. We plan to later replace this with a Drobo connected via FW800 so it’s easily expandable without me, and keep this OpenFiler server running just for storing backups to.

Next came the fun part of importing Exchange mailboxes. I prepared for this by exporting all the mailboxes out of Exchange 2003 into PST files. Little did I know that when it did this that it automatically password protected each PST. I didn’t find out until after about an hour or two of cursing at the Exchange 2007 Powershell. This LINK and this LINK helped with the syntax, but it wasn’t until I found this GEM that I finally regained my sanity and steamed ahead. Not only do you need to repackage the PST’s without a password, but it has to be repaired and checked so it is correctly imported. What a pain.

With that out of the way, we rejoined all of the workstations to the new domain with the help of SBS’s Connect wizard, which is much improved with SBS2008 I must say. It did a much better job of setting up all the settings for the users than that of 2003’s.

Finally there was the firewall upgrade. The aging harddrive was also replaced with a CF+IDE adapter combo for the OS. I dropped IPCop in favor of Smoothwall 3 because of it’s nicer interface and plugins. I loaded “FullFirewall Control” and “Zerina” plugins so I could have better firewall control and OpenVPN support. Users will be using the Microsoft VPN as part of Windows Server, but I wanted to use OpenVPN for myself. The goal of OpenVPN was so in the event the Dell server was down (thus VPN was down), I could still plug into the office network remotely and connect to the DRAC (Dell Remote Access Card) on the server and see what it’s hardware status was or what was on the console screen. We ran into a problem with this right after I left though (figures). I had forgotten to change the DNS server entries in the firewall config to those of OpenDNS, which shouldn’t be a big deal to change. But once I changed this, the firewall went down, taking the office with it. I walked someone through changing the config back from the console but it didn’t help. We were baffled by this, such a change shouldn’t have affected anything. Luckily I had a small Watchguard SOHO there for just an occasion, so that got them up and running until I could figure out the cause. Unfortunately, after a week of testing, I couldn’t recreate the problem but I did managed to get it working again. It is now fully functional, OpenVPN and all.

Aside from a few minor issues of getting files restored and profiles setup correctly, everything worked out well, although the entire process took more time than I had originally planned. To date, it has been one of the smoothest upgrades we have ever done, and SBS2008 has worked out really well. Working with SBS2008 is worlds better than 2003 was, mostly in part to the combined technologies of Server 2008 and Home Server, both great products on their own.

Tidbits:

Here are a few extra bits I came across that might be of use to others.

We use Gmail as our frontend server to filter incoming email, and act as a backup spooler of sorts. SBS2008 “finally” adds support for SSL POP3 retrieval natively, so that was easy to setup. In 2003, I had to setup OpenSSL and use it as a local proxy Exchange could use to connect to Gmail through since it didn’t support SSL natively. Every 10 mins Exchange downloads email from Google Hosted Domain Gmail accounts into the appropriate Exchange mailbox’s. Having Verbose Logging enabled helps too. For us though, only our office staff uses Exchange mailboxes, and our remote staff only uses a Gmail account that we provide. So, somehow we have to get Exchange to route local email to Gmail for those users who don’t have a mailbox in Exchange. This is done by creating a custom Send Connector that sends any Unresolved Recipients up to Gmail’s servers for routing. For any other emails, those get routed via the standard Send Connector to our hosting server that acts as a smarthost.

We also have some users (like me) who only use Gmail for all email and want to send their company specific emails directly through the Exchange server instead of to Gmail and wait for them to be downloaded by Exchange. The easiest way was to enable port 587 for Gmail to connect to as part of its new Send Mail Through ability. This just basically requires a custom Receive Connector as documented HERE.

Other great links:

http://blogs.technet.com/sbs/archive/2008/06/03/sbs-2008-how-to-info-on-david-overton-s-blog.aspx

http://sbs.editme.com/

The Home Server Specs:
Intel Atom 330
2GB Ram
Firewire card
LianLi PC-Q7 Case

In total, this was about $190, I also got a replacement 40mm fan for the motherboard as the reviews said it could be noisy, which was a good idea it turns out later. I had a spare laptop SATA drive lying around along with a 1TB 3.5 inch SATA drive, both of which I put in the case. Same goes for a 500 Watt PSU and CDROM, except the CDROM was only to load Windows, I took it out later after the install. The final install was a bit cramped only because the case can take a full size PSU (and all of it’s cables), but other than that, I love the case. It’s the perfect size for sitting on our bookshelf with the switches and modem. Speaking of switches, I opted to get a gigabit switch off ebay and since I have the venerable Linksys WRT54GL (Hacked with DD-WRT and bigger antennas). I found the older model Linksys EG008W so they stack nicely, costed 20 bucks. After I got some short Cat5 patch cables I got it all hooked up I noticed the fan in the gigabit switch (yes, it’s got  a fan!) was noisy and grinding. I popped it open and low and behold, it’s a 40mm fan! So I dropped in the stock fan from the Atom motherboard as an easy fix, runs fairly quiet now.

Firewire?? Yes, the Atom board doesn’t have onboard Firewire but I still use an OWC Mercury Elite Pro dual drive firewire 400/800 enclosure I got a while ago that houses two 500gig IDE drives. All my data is on these so I opted to keep using it. I am currently moving all of my data to the internal 1TB drive, then I will use one 500gig drive as backup drive for the system, then add the other to the drive pool for share duplication. This has the added benefit of letting me unplug it in a hurry and having all of my data and server data if I need to restore it. My extended plan is to eventually get a Drobo and just fill it with 1TB drives BUT I wanted it connected over FW800 so I could get the best throughput. USB2 is limited to about 480Mb/s which is abysmal compared to FW800’s 800Mb/s. It’s not eSATA speed but I can still switch to that later if I want.

In a past post I mentioned I use a series of online services for backups. Now that I have a central server, I am going to move to using iDrive to backup just the server data offsite and stop using it on my desktop. That move will include redirecting my “My Documents” folder to the Home Server so everything is centralized. Now I know there are a few add-ins for Home Server to backup to the cloud, like KeepVault or JungleDisk, but I am still not as happy with either as much as I am with iDrive, so far..

As for streaming and accessing the Home Server remotely, I have opted to keep using Orb because I like how it can handle encoding/transcoding to either my 360 or my iphone; and I have been using the free LogMeIn service for basic remote access to the desktop. I think this keeps things a little more secure and keeps from having to punch holes in my firewall (or worse UPnP) to grant remote access. I also have DynDNS running on my router so my address is always up to date allowing me remote access to the home server’s webpage. So far so good, I haven’t run into any major issues other than not scheduling iDrive correctly which let it suck up all my bandwidth all day as it ran its backup. Oops. I’ll come back later and writeup a long distance test drive review.

At my side-job, where I moonlight as a hosting network admin, we are preparing to migrate most of our clients to Rackspace’s CloudSites. I was very pleased with Rackspace’s performance and options while testing. Compared to the competition, they offered the best options and easiest to manage system. It’s still growing but so far I am impressed. For $100 bucks a month, we can run most of our hosting network in the Cloud compared to the almost $900 a month spent on a beefy VMware server and dedicated CDP server. While the setup has worked out great, I don’t have as much time to babysit it, nor does the company have the cashflow to support it in this economy. It will also be nice for me to not have to worry about patching vulnerabilities, server updates, or spikes in traffic. I also really want to get away from Plesk. While I still like it compared to other control panels, there are just quirks about it that should have been fixed long ago, and it’s just another yearly cost in licensing. Of course, moving to the cloud has presented it’s own set of issues such as needing to overhaul our development process and environment, update old websites to work correctly in the new environment, and setup permanent solutions to old duck-taped ones. I will try to remember to post what I did when the dust settles.

Over the last few months I have been working in my spare time on a plan to upgrade the office server for said side-job company. The current server is about four years old and is beginning to show. The hardware is decent, but it’s still all patchwork on a shoe-string budget, i.e. enough to get by as a startup. It runs Small business Server 2003, but it’s had so many fixes and changes made to it that it’s a bear to manage anymore. Now that the company is hitting it’s stride (and then the market puked), it’s time to set them up with something a little more robust and make my life easier. We retired some of our Dell servers at work and put them up on auction, one of which I won. It’s a Dell Poweredge 2850 but with a couple upgrades and spare parts from a dead 2850, I was able to beef it up quite a bit. I also grabbed a DRAC (Dell Remote Access Card) off ebay for cheap to drop in so I finally have remote console ability (I’ve dreamed of this for years). I plan on installing Small Business Server 2008 along with upgrading an aging storage server install to Windows Home Server, with some tweaks. This is not only a performance upgrade, but it’s a clean slate with Exchange 2007 and Sharepoint built in that will be greatly welcomed.

I won’t actually be using the features of WHS other than it’s Drive Extender functionality for the office shared folders. It’s currently got a RAID10 for 4×250GB drives. It does well enough performance wise but the actual server itself is aging. If it’s in the cards later, I’d prefer they just move everything to a Drobo and be done with it (over Firewire) and retire this server. I’ve dropped a spare FW card into the Dell for just an occasion.

The next major upgrade piece is the firewall server. It’s had a few hiccups this last year so I want to remove it’s old harddrive and replace it with a CF>IDE combo. That should improve performance and reliability considerably. I will be dropping IPCop in favor of SmoothWall, mainly for the better control panel for traffic monitoring and OpenVPN addon. While they use MS VPN built into Windows, this only works if the office SBS server is up and running. Utilizing the OpenVPN setup, I can remote into the network independently and connect to the DRAC in the Dell to troubleshoot. I’ve already setup the entire office environment inside of VMWare here at home and it works really well.

All that remains is to get all of the desktops prepared and export everyone’s current Exchange account. We will be doing a completely clean install of the new system instead of a migration so we don’t bring over any old cruft. This should be interesting.

As of this week though, I finally got to try out the new “XP Mode” for backwards compatibly. Overall I like its implementation, but there are a few things I think they should change before final release. When you really examine it, XP Mode is really no different than running the other virtualization packages like VMWare Workstation/Fusion or Parallels Desktop. XP Mode is really Virtual PC running an XP virtual machine (a full OS in a self contained software machine) sprinkled with Microsofts Med-V virtualization running applications in a “coherence” sort of way to give the impression your running the application natively, but in actuality the XP desktop is just hidden. In other words, nothing groundbreaking.

Applications installed within XP Mode show up in your start menu in Windows 7, and when clicked, kick off (boot up) the XP virtual machine and run the application. The Windows XP desktop is hidden, but when you move your application window around, you can see it bleed through around the edges as it redraws the screen. VMWare Fusion does the same thing on my Mac with Windows in “Unity” mode. Not a big deal, but not quite “seamless”. The ramifications of this can be huge though since this now allows Microsoft to scrap all sorts of legacy code and rely on XP Mode to take care of any issues. This will severely help lean out Windows and the years of code carried between version to make corporations happy.

There are a couple things I would like to see changed based on my first impressions though:

1. I would like to see the ability to choose where to install an application. From within Windows 7 with XP Mode installed, if I start an application installer, I would like Windows 7 to ask if I want to install it natively, or into XP Mode. Instead, right now I have to startup XP Mode and go into the desktop of XP and install an application before it is displayed in my Windows 7 Start menu.

2. I haven’t tried too many applications yet (like Office) but the interaction with the native system (that being Windows 7) sort of works. If I download or try to save a file, the default location is my native desktop. If you try a different location though, you have to navigate your way through to your native drive that is “mapped” as if it were a network drive. In this case, I see a “Local Disk C:” which is XP, and a “C ON Windows 7 system”. Not the easiest way to go about this.

Those were the only major faults I could find up front. Keep in mind though, since it’s basically Virtual PC, you can tweak the settings for XP to give it more ram, more drive space, etc. You can even opt to install a different OS like your flavor of Linux. For now, I am only running this on my Macbook pro via BootCamp. I haven’t brought myself to wipe out my Vista desktop and install 7 yet. I also need to do some more digging to compare it to my workhorse VMWare Workstation where I spend most of my time. To be honest, nothing I run has had a problem with Windows 7 or Vista, so XP Mode might be a lose on me. Using it in corporate environments probably wouldn’t work either, but that’s where App-V and Med-V come into play. We shall see…

The first place is to a second set of 500GB drives in a Firewire enclosure, one drive for a mirror, the other for drive images of the system drive. To do the mirroring, I use a program called Mirror Folder, which I have had forever. It does folder watching and either real-time sycning, or scheduled syncing between a source and destination. Everything on the storage drive gets mirrored every two hours to one of the Firewire drives. In addition to this, I use Acronis True Image 2009 to make an image of my system partition (the dual 640GB drives) twice a month that is stored on the second Firewire drive. This way, if the house is burning down, I just grab the Firewire drive and run. This will give me a usable image of my system, as well as a backup of all of my archived data.

If that fails, I have a secondary backup using IDrive online backup service. Every night I have a set of watched folders (the most important stuff) backed up to their service online that I can restore to any machine, and access via their website if I ever need to. They also keep revisions of files as they change, letting me roll back if I botch something horribly. It will not restore my whole machine or anything, but it gives me piece of mind knowing my most important data is backed up somewhere offsite. It came down between Carbonite and IDrive and I ended up choosing IDrive for its better usability both online and via its desktop application. It’s not unlimited space like Carbonite, but I’m not using all of the space anyway, so no big deal.

Now, this is just my backup process for worst case disaster scenarios. For daily syncing/sharing of files between the multitude of computers that I use, I primarily use DropBox to sync files, and LiveDrive for online sharing and storing of random stuff. I recently switched from ZumoDrive to LiveDrive because they allow uploads within their web interface. My other reason for switching was since I was part of their beta, they offered me a one year unlimited account for half off. I figured I would give it a shot and let everyone else catch up and re-evaluate them all at the end of the year. I am a HUGE fan of DropBox, so hopefully by then, they will have a few more key features added so I can rely just on them (and better pricing).

So there you have it.

I also don’t get why everyone keeps criticizing Apple for not having one, or spreading rumors that they will have one at some point. Why? Apple already has two notebooks that fit in the category of “sub-notebook”, granted, they are a bit bigger than the normal Netbooks, but they are still small, and far more functional. I’m talking about the $999 13″ Macbook, based on the previous white model, and the super thin Macbook Air. While they are out of the price range of typical Netbooks, but they are far more functional, and have full size screens and keyboards. I would much rather spend the extra money one something I can get plenty of use out of as opposed to a cheap product that I can only do a few things on, then switch to a different product to do the rest. But I guess that’s just me. And we are talking about Apple here, if there isn’t a decent profit margin to be had, then it’s not worth their time and development. If they would fix some of the glaring omissions in the iPhone, they would have no need for a Netbook.