Tom Petrocelli's take on technology. Tom is the author of the book "Data Protection and Information Lifecycle Management" and a natural technology curmudgeon. This blog represents only my own views and not those of my employer, Enterprise Strategy Group. Frankly, mine are more amusing.

Tuesday, September 29, 2009

Of Paperweights and Doorstops

I have quite the collection of paperweights and doorstops. They didn't start out that way of course. They all began life as usable electronics like cell phones and networking gear. All were the victims of upgrades to newer gear such as Wireless-B to Wireless-G. My cell phone upgrades every two years per the plan, leaving a graveyard of old phones. Some devices were on their way to the trash and I couldn't help but save them from becoming toxic waste. Don't kid yourself. Even the most innocuous electronic devices have heavy metals, plastics, and other materials that will continue to pollute for centuries to come.

What annoys me is not the proliferation of electronic gear but the fact that they can't easily be reused or upgraded. Why? Because they are designed to be tossed away, use closed architectures or, and this really galls me, are purposefully locked.

For example, I decided to upgrade my cell phone this year. My old one worked fine but was lousy for texting. So I took the old one to a Verizon store and asked if they could convert it to a Tracfone. Tracfone is a pay-as-you-go cellular service provider which is great for the kids. The no-contract, phone card type system keeps teenagers from running up massive phone bills. I was told, quite emphatically, “No!” What's so ironic about this is that Verizon owns Tracfone. I was offering to spend more money with them but they didn't want it. So the phone sits in my desk draw waiting for its day of liberation.

I could attempt to jailbreak it. Jailbreaking consists of hacking the hardware and firmware to remove whatever is tying the device to the service provider. Why should I have to do that though? Besides the fact that I might brick the device, it's also likely that Tracfone/Verizon still won't let me use it.

There are alternatives. I could donate the phones to various groups that re-purpose them for soldiers in Iraq or victims of domestic violence. Worthy use but I wanted to get more out these myself. And I'm pretty sure they don't want an old DSL Modem.

The sheer waste is incredible. Millions of these devices get tossed in landfills or, hopefully, sent to a recycler. All could be given a new lease on life if only they could be opened up, added to, and tinkered with. I'm not suggesting that vendors open up the phones when the are active. That would be nice but unnecessary. It's great that Cisco opened up their Linksys Routers. Lots of hackers enjoy extending their WRT54 devices, adding new features and sometime using them for entirely different purposes. I'm not that ambitious. I only want my devices to have a longer lifespan.

One of the great things about computers is that they can be used nearly forever. I know a lot of folks that still use DOS era computers for useful purposes. Some are hobbyists and others use them for a single purpose like voice mail. I still use a nine year old 40GB disk. I stuck it in a USB drive case and use it for email backup. Why can't we do that with all of these devices? Okay, it's big company greed but it's stupid greed. They could sell me a cheap retrofit kit and make a couple of bucks that they wouldn't have.

Reuse is the ultimate recycle. Let my devices go!

Wednesday, September 23, 2009

Rollups In Our Lunchbox

There's been a lot of chatter about the acquisition of Perot Systems by Dell this week. It's not surprising that in a slow news week so much attention is being given to a deal that is only one of many that have happened in the past few months. Most of the commentary is the typical stuff such as “This will (or will not) help Dell compete against IBM and HP.” or “Dell service partners will (or will not) be mad as hell.”

I'm more interested in what the Perot acquisition says about the state of the computer industry. Since 2000 the computer industry has been consolidating, coalescing into a handful of companies that directly deliver solutions to customers. The rest of the industry exists to serve these mega companies. It's like the car industry. There are only a half dozen companies that sell cars and services to the vast majority of consumers. The rest supply parts such as headlights and seats.

You can easily tick off the number of computer companies that sell the majority of equipment to both business and consumers. Dell, HP, IBM, Cisco, and EMC are delivering complete solution sets in all segments including enterprise computing of all types, consulting services, and consumer products (through brands like Linksys and Iomega). Microsoft is a special case. It is a major, first tier company that sells mostly software (mice and Xbox systems aside). I wonder how long it will be before Ballmer buys a large or specialty hardware company like Acer. I wonder about the same for EMC. How long before they buy a specialty server company or small networking company.

There are a few companies overseas that are part of the mix as well such as Hitachi in Japan, the aforementioned Acer in Taiwan, Lenovo and Huawei in China and Groupe Bull in Europe but their sales, product depth, and international reach pale in comparison to the biggest five or six American companies. Don't kid yourself though. Any of those could be the next Toyota. Or a great acquisition. These folks compare to Fiat. Not GM or Ford but certainly an important company.

Apple is important but still small in comparison. They are like Porsche - a well-known, high-end luxury brand that people pay more for because they want something they perceive of as special. They are technology trailblazers with a great sense of style. More taste maker than mainstream sales giant.

Unlike the auto industry, there is still a healthy second tier that is trying to move up. NetApp comes to mind. Brocade and Juniper too. These companies are centered around being best of breed in one area of the industry such as data storage or optical network equipment. This segment keeps getting smaller though. How much longer can Brocade or Quantum remain independent? Sooner or later the combination of size and narrow focus will make competing against the top tier impossible. They won't be able to offer anything unique enough or broad enough. They will either sell out in some face saving “merger” or go the way of SGI, becoming mere shadows of themselves. Perot Systems is a great example of a second tier player (IBM GS and HP Consulting they are not) being absorbed by the first tier. Second tier players provide entry into a whole new segment for a big company.

There is also a whole host of small companies that supply specialty equipment or develop new technologies for big companies. These small companies are technology Spackle. They fill holes in a big company's product line. A few will grow independently but most will be gobbled up by bigger companies. Some will fail altogether. A great many will stay small, too small to acquire. They will continue to survive by supplying critical parts to the big boys. There are lots of small companies that make specialty chip designs or cores, RAID controllers, and software libraries.

This year has been a banner year for rolling up small companies. With credit tight and VC money still not flowing free a great many small companies are realizing that, despite current success, they won't find the money to grow. Really, how many computer technology IPOs are going to happen over the next two years? For many small companies it's grow or die... or sell out. For the big companies this is a boon if they have cash on hand. They can buy out the smaller players for much less money then they can in good times.

So, what's next. I think the majors will continue to buy companies. Dlink is a good acquisition target. They have a great consumer and SOHO line that would help out HP or Dell. The macro trend will continue for awhile. Tight credit and investor wariness will make life difficult on the companies exiting the startup phase. The biggest companies will take this opportunity to reinforce or extend their product lines, perhaps even get into new segments.

So, if you aren't selling cars to millions of people, you are probably selling headlights to the guys who make the cars. Don't feel bad. There's opportunity in that.

Monday, September 14, 2009

Free Software Tools for Geeks

Everyone knows that techies have different needs when it comes to software. Come on. Admit it! For the average person a slow network is a mystery and an annoyance. For us, it's a project and all projects need tools.

Over the years, I have accumulated many software tools that can be had for free. Some are open source tools supported by a vast community of developers. Others are a hobbyist's pride, given freely for all to enjoy. Some are particularly useful.

So here is my top software tools for the serious computer geek, with commentary. Did you expect any less?

PuTTY

If you don't occasionally need a terminal session then you have no right to consider yourself a computer geek. The command line is what separates the real deal from the poser. PuTTY is an excellent Telnet interface with support for SSH, Rlogin, and even a serial shell.

Pros

Git 'er done! PuTTY gets you a command line to almost anything. It emulates the most common types of terminal (ah! The VT100. You never forget your first) and has a bazillion options to tweak your terminal session. Mostly, the default settings are all you need.

Cons

PuTTY only handles one session at a time. You have to load multiple instances of the software to talk to multiple systems. Even more annoying, when you shut down a session, the whole program shuts down and you have to reload it to talk to another box. Still, these are annoyances not major flaws.

KeePass

There are lots of websites to visit, servers to manage, and PCs at home and in the office. All have passwords to manage and it's a pain. If you use the same password over and over you know that's a security risk. Besides, the user name and password requirements vary from site to site and box to box. Keeping an unencrypted file or database of passwords on your computer is inviting disaster.

Enter KeePass. It manages an encrypted database of information about your logins. Besides storing user names and passwords, KeePass also has search and organization capabilities.

Pros

Does the basic job of storing user names and passwords extremely well. Search is fast and accurate. The password generator is also useful when you want to create secure but different passwords.

Cons

The interface is a bit dull but, really, it's not a game after all.

MailStore Home

Unlike most people, techies tend to generate and receive a lot of emails. No matter how good your email system, sooner or later your system bogs down if you don't archive them. Or you bog down trying to find needles in your email haystack.

MailStore Home is the little brother, freeware version of a commercial package. It does a credible job of archiving and indexing emails. I find it useful for archiving emails to a USB drive which acts as a backup. I can then delete emails from my email client with confidence.

Pros

MailStore Home interfaces with most of the major clients such as Outlook and Thunderbird but also can archive from a server, including Exchange, POP3, and IMAP mailboxes, and webmail systems such as Google Mail. It's also pretty fast for an average techie which means an above average email user.

Cons

It can copy but not move emails. That's great if you want to backup your email but not so great if you want to truly archive them. Instead, you have to remember to go back to your email client or server and delete emails manually.

Sun VirtualBox

In the world of virtualization, VMWare has the mind share and Microsoft's VirtualPC comes bundled with their servers. Sun's VirtualBox is not as well known which is too bad. It's best feature is that it is really easy to use. You can run most anything you want with minimal effort. It's free for individual use which makes it a great choice for home or a hobbyist. Would I run a data center cloud on it? Probably not. For testing, developing, or just plain goofing around, it's so much easier to use.

Pros

Easy to use. You don't need a four week course to start using it. It does a very credible job of creating virtual servers or desktops.

Cons

Configuring inbound network access, such as an HTTP server, is not intuitive making VirtualBox more useful for virtual desktops or sandboxes. I still can't get FreeNAS to work right because of server access problems.

The list of freebie tools is much longer than this. This is but a sample of tools for the techie. Also in the mix is 7-Zip, an excellent archiver, and Filezilla, a classic and profoundly useful FTP client. For the software developer, I also recommend Sun's Netbeans. It's a full blown, commercial quality, IDE that is especially good for Java development but has decent support for PHP, Python, C/C++, and many other languages.

A lot of these tools are much better than the stuff you pay for. Some of them you do have to pay for if you want to use them in a commercial setting. It's always a good idea to check the license. And, if you are managing a large commercial environment, many of the tools won't provide the scope of features and services that you need. However, for the hobbyist, individual, or SOHO environment, these tools can't be beat. They give you what you need for a great price – free!

Friday, September 11, 2009

eyeSpy With My Eye My Desktop In The Sky

I've been intrigued with the idea of server-based desktops for awhile. In the past I have toyed around with Desktop Two and the recently deceased Jooce. While interesting, they all suffered from the same set of problems. They were usually:

  • slow;

  • lacking in useful applications, and;

  • had significant security problems.

The last item is significant. Despite what the folks at Google would tell you, most companies don't want to keep sensitive files on someone else's server. The most standout issue with virtual, online desktops (which, heaven help me, they are calling Cloud Desktops) is that there didn't seem to be enough need for them. Until now. I'll get to that later.

eyeOS overcomes the majority of cloud desktop problems. Written in PHP5, you can install it on your own internal server or even a web hosting server. It's open source to boot! This is important since it makes it easy (or even possible) to create your own web-based applications to run in the virtual desktops. The suite of applications that comes with eyeOS is pretty good but the ability to create your own makes it a much more useful platform for business.

Being able to install eyeOS on the server of your choice also goes a long way to removing security objections as well. IT can now control the security environment and does not have to rely on blind faith in a service provider. The same goes for performance. You can use your own magic and hardware to up performance levels to where you want them to be. Even installed on a web hosting platform, eyeOS had a decent response time, granted for a small number of users. That it worked at all in anything resembling a useful manner was pretty startling.

The big question that is still unanswered is “who cares?” That's the acid test for all products. Why should I spend any time at all, let alone money, on this product. As little as two years ago, I would have to have answered “ I don't care.” Today, the world is a different place. We now have a proliferation of small, Internet enabled devices including smartphones and netbooks. Many of these are too underpowered to have a full range of desktop applications – or do they?

This is why eyeOS really makes sense now. Virtual desktops have been primarily concerned with giving users a consistent and controlled desktop experience throughout an Enterprise network. That's fine except that it doesn't carry on to other devices. With eyeOS, you can set up your own cloud desktop service, that meets your standards for security, and make it available on anything with a browser.

With eyeOS in hand, you can outfit your sales force with cheap netbooks and still give them a full range of office and corporate applications. In other words, the whole desktop experience. For an added bonus, if users stick with cloud desktop for everything, you don't lose or expose your data even if they lose their PDA, smartphone, or netbook. It's not on those devices, it's on the server.

The best part is that they users don't have to do anything special to make this work. They sign into the eyeOS server and away they go. Users don't have to remember to copy files to encrypted drives or anything like that. They just do what they normally would do on a desktop computer. In terms of data security, this is a great leap forward. And the fact that only administrators can install applications is sure to please corporate security types. Fewer rogue applications in the corporate network is a good thing indeed.

eyeOS is not perfect by any means. Many of the critical applications, email especially, are nowhere near what a decent corporate application should be. However, it is clear from the Zoho widgets (downloaded separately) that you can integrate other online applications into eyeOS. With more and more companies going to web-enabled applications anyway, lack of sophisticated, standalone desktop applications is really not a problem.

Cloud desktops and eyeOS in particular, are not quite there yet. However, they are rapidly getting there. An organization that is committed to cloud desktops could make eyeOS into what it needed. Not out of the box of course but with a relatively small amount of effort.

The old fashioned, fat desktop will never go away. There are too many applications that will never port to a platform like this. I don't see programmers writing serious code on a cloud desktop. For the average wage slave, however, this would be an improvement and IT will love it.

Thursday, September 03, 2009

EMC Pacmans Kazeon. Mmmm. Good.

More news on the storage M&A front and once again it's about EMC. EMC announced that they are acquiring Kazeon a self avowed “leader in eDiscovery” tools. Stripped of all the eDiscovery hoopla, Kazeon makes a decent indexing and classification rules engine. In that space, it is a very useful thing indeed.

What Kazeon does is catalog and index your data and run the information past a rules engine to generate additional metadata. Again, that's good but only as good as the rules. They also layer on some analytics which is all the rage in eDiscovery. Analytics are also only as good as the data and metadata and, in my opinion, overemphasized in the eDiscovery market. But that's just me...

Kazeon is a hand in glove fit for EMC. For many years now EMC has looked to get past selling just hardware and has wanted to sell systems to store and manage data. That's a great strategy since it creates new value but sticks to their knitting. Kazeon is another value added tool that EMC can add to their toolbox.

The Kazeon acquisition also gives them some eDiscovery street cred. They have been trying to play in the eDiscovery sandbox for years, mostly through their Documentum offerings. Nothing wrong with that since the majority of eDiscovery work is about workflows anyway. However, automated tools for tracking data also are important not only to discovery during litigation but also to insure compliance with in-house data retention rules. And by retention I mean destruction. But you knew that didn't you...

The dirty secret about data retention is that no one can really comply with their own internal rules without knowing where and what all their data is. Knowing where all your data is is a really hard problem to solve. That's where Kazeon comes in. They create catalogs of vast amounts of data that allows you to better comply with discovery rules, a preservation and legal hold, and internal data retention policies.

So, Kazeon is an obviously good thing but why is it good for EMC? Actually, there are two (probably more) reasons why this works so well. First, it adds value. If I buy tons of EMC storage, the Kazeon/EMC products will help me to know what I have on it. Second, those catalogs of information and metadata need, you guessed it, more storage. It's the reason Documentum was such a good deal for EMC. It lets you get more value from your stored data and makes you store more data. A twofer for EMC.

EMC will now be able to deliver, by itself, one of the most comprehensive information management systems available. By combining Documentum, Kazeon, and all the other tools EMC has at its disposal, plus hardware, they will be able to deliver an information management solution that will make lawyers squeal with delight.

That's not to say it's perfect. Kazeon can't help you if someone dumps their files onto a flash drive or emails a smoking gun document to their Gmail account. Smartphones and PDAs are also a challenge that Kazeon will not help with. Still, they hit all the high notes and do better than what most companies do - which is nothing!

As an aside, Kazeon also has an intellectual property (IP) management component to their systems. IP management and eDiscovery are very similar problems – you need to know what data you have where, in order to comply with rules and regulations. EMC has often touted Documentum as an IP management tool. They haven't gotten too far with that since it takes so much effort to set up Documentum to do IP management. Unless you are already committed to Documentum across the company, there are better out of the box IP solutions. Kazeon will give them some more heft in that space. It will allow EMC to automate many of the sticky tracking and classification tasks associated with IP management, especially in preventing leakage. It's not there yet, but getting better.

I don't know if EMC is full yet after eating up so many companies. Kazeon is quite a tasty and healthy morsel for them though. It makes good, strategic sense. I wonder if they left room for dessert.