There are companies such as Picky Domain Names that specialize in finding the perfect URL for your business, but there is no reason why you can’t do the same thing yourself, for free! All it takes is a good imagination and some time to brainstorm. If you have friends who can help, all the better.

Most of the regular domain names have been bought up and used now, so it is a real challenge to find ones that are unique, describe your business and are easy to remember. That’s why you need to get creative.

-First off, there are a few general ground rules to ensure that your visitors remember your domain names. This is obviously the most important part of choosing an URL, so keep the following in mind:

-Limit yourself to a maximum of three words. Longer phrases tend to be difficult to remember and also they take too long to type in. People want information fast and if they have to try and remember a twenty letter URL, they probably won’t bother.

-Use words that are easy to spell. If you use weirdly spelled words just to get the domain names you want, you will lose a lot of visitors. For example, if myspace.com is taken, don’t go for myspayce.com, you will just end up sending traffic to the other domain name.

-Think carefully before using anything apart from .com as an ending for your domain names. It is the most common and often people will forget if you use a more unusual ending such as .name or .tv.

Now that you have the basics in mind, it is time to brainstorm. Start by writing down all the words you can think of that relate to your business. For example, if you have a diaper service, you would end up with a list of keywords that include “diaper, diapers, baby, bottom, cloth, wet, dirty, clean, service” etc. Include any slang or nicknames as well, such as “nappy, stinky,” etc.

Once you have a good long list, start trying to put the words together in a form that would work as domain names. You should be able to come up with at least a dozen that fit the description of your business. For our diaper service example you might end up with nomorestink.com, dirtytoclean.com, nappybottom.com, etc.

Next, you need to go to a domain names registrar such as GoDaddy and check to see if your domain names are available. Start with the ones you like best and work your way down the list. Chances are you will find at least two or three that are available. These are then your choices for domain names!

This whole process works best if you have two or more people to bounce ideas off of. Domain names that are wacky and outrageous are sure to come up as you work your way through a brainstorming session with friends or coworkers! And at the end of it all, you should have a handful of domain names that you can go with.

• 0 Comments

It’s another day on the Internet and yet another form of ad monetization has reared its ugly head.

You Tube recently launched its InVideo advertising solution to a resounding dull clang of dissatisfaction from the exuberant masses who consume hours of content on the beloved Web site.

It’s not hard to see why. And in the interests of full disclosure, I do work in online advertising, yet I think it’s really kind of hard to love this solution.

Don’t get me wrong. For some, loving advertising is like loving tax preparation. YouTube’s solution isn’t easy to understand in the context of how it’s used with content. The relative context of application, if you will.

Specifically, hearing a lo-fi video defense of a pop star is bad enough, but when it’s shrouded with a slick transition to the ad message, it all seems a brash and out-of-place marriage of elements.

This blight-on-blight online advertising solution is only the most recent. There have been many others before, and rest assured there will be more in the future.

It makes me wonder why we keep coming up with the same solutions to monetize this relatively new medium. Can’t we get a little more creative? Or are we destined to a downward spiral of repeating the same advertising model on everything we do?

Make the logo bigger. Stick a special offer in at the end. Repeat. Yes, these forms of advertising are familiar, like a comfy old chair, but old and comfy can get stinky after a while.

In their book, “New World, New Mind” Robert Ornstein and Paul Ehrlich point out, “Human judgment and humanity’s ability to deal with the consequences of its creations lags behind its ability to create.”

Bad times at any company’s online solutions division lie ahead.

We constantly trying to innovate, yet we don’t slow down our minds (or our development cycles) to really contemplate the impact of our work. It’s a truly vicious circle.

We develop measurement criteria in the same way. Often, these two crimes combine to provide poor ad solutions and mishandled performance valuations.

In a competitive industry, yes, time is money. But ignorance wastes time, money, and a user’s interest in your brand. I’d say the last is much more costly.

Damning as that may sound, it’s been true for quite a while. Did anybody ever think the floating rich media ad would dominate the online advertising industry forever?

For that matter, the lovely and controversial You-Tubian solution for integrating advertising into its spectacle of video content shares the same future.

Lest we forget, the practice of inserting :15 spots before and after video content that floods the news became the new hair extension for a thinning and expensive media market. Truly, to think an online spot can be included in overall television rating points (TRPs), regardless of screen size and format, is unfathomable.

But as bad as it sounds, there’s hope for an industry known for interrupting viewers’ valuable time (and being the spawn of Satan).

There are ways in which we can rethink this whole ad thing. First, we must ask ourselves a few questions: Why must an ad follow a traditional story arc? Why do we need to assume users must understand everything they see? Why do we need comedy to make a viewer remember an ad? Why can’t the user have several available options and interactions?

—

By  Dorian Sweet

• 0 Comments

Deutsche Telekom’s T-Mobile will have exclusive rights to distribute the Apple iPhone in Germany. The Germans will not be happy with the price tag of owning an iPhone: 399 Euros or about $554. I suppose the rumored T-Mobile ad was slightly off by 100 Euros. It remains to be seen whether Germany will line up for the €399 or wait for the price cut. Being Western Europeans who are used to paying premium for goods and services, we suspect they will line up.

 

• 0 Comments

In a big first, YouTube has decided to let advertisers inject their messages inside the video frame for select content on its site. The new offering, dubbed InVideo Ads, mimics the clickable ad overlays introduced in recent months on ad networks like VideoEgg and YuMe.

In testing for several weeks, the ad product consists of animated bars that obscure the bottom 20 percent of the video frame for a given clip. They initiate 15 seconds after the beginning of a clip, an interval YouTube selected because “it takes users 10 seconds to become acclimated to what they are watching,” according to Shiva Rajaraman, YouTube Product Manager. InVideo overlays are “80 percent transparent” and remain visible for approximately 10 seconds before shrinking to a small button users can later click to view the marketing message again.

YouTube has set a $20 CPM for InVideo ad buys consisting of an InVideo ad accompanied by a tiny in-player companion ad and an adjacent in-page unit. Brands advertised in the testing phase include Universal Studios’ “Evan Almighty,” Twentieth Century Fox’s “The Simpsons Movie,” BMW’s 3 Series convertible, NewLine’s “Hairspray” and approximately 15 others. Select content providers taking part in InVideo — and in any ad revenue it generates — at launch include Ford Models, Warner Music Group and Roadrunner Records. Several of the content examples YouTube shared during a press briefing consist of ads placed within music videos.

Rajaraman said the field of participating content owners is “significant,” but declined to identify their exact number or to estimate the volume of ad impressions YouTube expects to offer media buyers under the new program. But he insisted reach is not a concern.

“Our main goal is to support a range of genres and advertisers, and frankly to support any size campaign,” he said. “Given the inventory there, we’re comfortable with supporting large as well as highly targeted campaigns.”

Clicking on an overlay ad pauses the current video and launches one of two experiences brands can choose between. One is a new clip superimposed over the video in progress via a player-within-a-player interface. When the paid clip ends or is closed, the original automatically picks up where it left off. Rajaraman said 76 percent of those who click the overlay and watch the video ad viewed the entire trailer for NewLine’s “Hairspray.”

The other option is a Flash-based interactive experience in which the user is invited to navigate an interactive menu. Warner Bros. created such a unit where users can flip through selected album covers (click for example).

During YouTube’s research process, Rajaraman said, “One of the key things we found, not surprisingly, is that when a video is playing on YouTube their attention is [locked in to the video frame]. When we came up with an ad format, we realized that… it needs to be in the player.”

Yet when the Google-owned video portal tested pre-roll placements, YouTube users abandoned video clips at a more than 50 percent rate. The overlay, by contrast, results in an abandonment rate under 10 percent. Not only that, but click rates are five to 10 times greater than standard display click-to-video ads, according to Rajaraman.

He added YouTube also gathered data during the test period by monitoring comments, speaking directly with some users, and observing navigation patterns.

YouTube says it serves 3 billion minutes of video each month, creating “a lot of opportunities for advertisers.” The vast majority of that content will not support its new InVideo ads, however, owing to the adjacency concerns so often raised by agencies on behalf of their most blight-sensitive brands. Those same concerns are no doubt also behind the YouTube decision not to insert ads in partner content that users choose to embed on other sites, such as blogs and MySpace pages.

—

By Zachary Rodgers

• 0 Comments

Leopard hacked beyond recognition and not even out yet…

Apple applauds Mac OS X as the most secure and most advanced operating system in the world. Microsoft in turn has adopted a more modest marketing approach, in spite of the Wow campaign, and is comparing Windows Vista mainly with Windows XP. In terms of security Vista is just the safest Windows version available and nothing more. The same argument is not valid for Apple. With the Cupertino-based company it is all or nothing, which means that it’s generally all. Just take the example of Safari 3 Beta browser for Windows. “Apple engineers designed Safari to be secure from day one,” stated a message posted on the Apple official website. The reality? Security researchers turn up no less than 18 security vulnerabilities for Safari 3 in the browser’s first day on 32-bit and 64-bit Windows Vista and Windows XP.

Charles Miller from Independent Security Evaluators has really made a name for himself after he managed to hack the iPhone. Still, the security researcher from Independent Security Evaluators seems to have an affinity for fresh Apple products. Or at least this is the direction underscored by his
presentation at Black Hat 2007 in Las Vegas: “Hacking Leopard: Tools and Techniques for Attacking the Newest Mac OS X.” Miller cites Apple with the following: “Mac OS X delivers the highest level of security through the adoption of industry standards, open software development and wise architectural decisions.”

Security was also Apple’s focus with Safari and iPhone and we have been able to see how well that turned up. But the upcoming release of Mac OS X, Leopard, scheduled for October 2007 is in a very poor condition in terms of security according to Miller. It is nothing but a joke compared to Windows Vista. And this is one aspect that emphasizes the differences between the next version of the world’s most attacked platform and an operating system with an inexistent threat environment.

Miller’s conclusion is worrisome for future Leopard users: “Why Hacking Macs is Easy.” According to Miller Macs are just as easy to hack as they are to use. “To help users, there are lots of 50+ suid root programs” revealed the security researcher. Suid Root is designed to help with the silent elevation of privileges in Unix and Unix based operating system such as the Mac OS X. Unix has had, a long time before Vista, access control capabilities, an equivalent to the User Account Control. Still, Suid Root is a design flaw, because allowing for silent and automatic elevation of privileges means inviting kernel level exploits.

In contrast, nothing similar to the Unix setuid/suid or sudo functionality can be found in the design of UAC in Vista. There is only one way that a service, application or process can gain elevation of privileges in Vista and that is only through the user.

Moreover, Apple does not “bother users with burdensome updates.” All the open source solutions included in Mac OS X are not kept up to date including OpenSSH, OpenSSL, Apache, Samba, Cups. “The Samba on Mac OS X had an exploitable remote root vulnerability in it…it hadn’t been updated since February 2005,” Miller stressed focusing on open source as an attack vector.

But of course there’s always the “safe from day one” Safari. Apple’s browser, and by the way version 3 is going by default into Leopard, launches the following programs on execution: “Address Book, Finder, iChat, Script Editor, iTunes, Dictionary, Help Viewer, iCal, Keynote, Mail, iPhoto, QuickTime Player, Sherlock, Terminal, BOMArchiveHelper, Preview and DiskImageMounter.” Any security vulnerability residing in any of these applications can be exploited via Safari.

In the end Miller exposes Apple security for the joke that it is saying that the company “makes exploitation fun,” mainly because creating exploits for Mac OS X is like going back in time to the software of 1999. The reason? “Apple doesn’t randomize anything: the location of the stack, the location of the heap, the location of the binary image, the location of dynamic libraries and (to top it all off) heap is executable.” By contrast Windows Vista introduces a security mitigation called Address Space Load Randomization (ASLR).

(source news.softpedia.com)

• 0 Comments

Broadband over power lines (BPL), also known as power-line internet or Powerband, is the use of PLC technology to provide broadband Internet access through ordinary power lines. A computer (or any other device) would need only to plug a BPL “modem” into any outlet in an equipped building to have high-speed Internet access.

BPL seems, at first glance, to offer benefits relative to regular cable or DSL connections: the extensive infrastructure already available would appear to allow people in remote locations to have access to the Internet with relatively little equipment investment by the utility. Also, such ubiquitous availability would make it much easier for other electronics, such as televisions or sound systems, to hook up.

However, variations in the physical characteristics of the electricity network and the current lack of IEEE standards mean that provisioning of the service is far from being a standardized, repeatable process, and the amount of bandwidth a BPL system can provide compared to cable and wireless is in question. Some industry observers believe the prospect of BPL will motivate DSL and cable operators to more quickly serve rural communities.

PLC modems transmit in medium and high frequency (1.6 to 30 MHz electric carrier). The asymmetric speed in the modem is generally from 256 kbit/s to 2.7 Mbit/s. In the repeater situated in the meter room the speed is up to 45 Mbit/s and can be connected to 256 PLC modems. In the medium voltage stations, the speed from the head ends to the Internet is up to 135 Mbit/s. To connect to the Internet, utilities can use optical fiber backbone or wireless link.

Differences in the electrical distribution systems in North America and Europe affect the implementation of BPL. In North America relatively few homes are connected to each distribution transformer, whereas European practice may have hundreds of homes connected to each substation. Since the BPL signals do not propagate through the distribution transformers, extra equipment is needed in the North American case. However, since bandwidth is limited this can increase the speed at which each household can connect, due to fewer people sharing the same line.

The system has a number of complex issues, the primary one being that power lines are inherently a very noisy environment. Every time a device turns on or off, it introduces a pop or click into the line. Energy-saving devices often introduce noisy harmonics into the line. The system must be designed to deal with these natural signaling disruptions and work around them.

Broadband over powerlines has developed faster in Europe than in the US due to a historical difference in power system design philosophies. Nearly all large power grids transmit power at high voltages in order to reduce transmission losses, then near the customer use step-down transformers to reduce the voltage. Since BPL signals cannot readily pass through transformers — their high inductance makes them act as low-pass filters, blocking high-frequency signals — repeaters must be attached to the transformers. In the US, it is common for a small transformer hung from a utility pole to service a single house. In Europe, it is more common for a somewhat larger transformer to service 10 or 100 houses. For delivering power to customers, this difference in design makes little difference with power distribution, but it means delivering BPL over the power grid of a typical US city will require an order of magnitude more repeaters than would be required in a comparable European city. One possible alternative is to use BPL as the backhaul for wireless communications, by for instance hanging Wi-Fi access points or cellphone base stations on utility poles, thus allowing end-users within a certain range to connect with equipment they already have. In the near future, BPL might also be used as a backhaul for WiMAX networks.

The second major issue is signal strength and operating frequency. The system is expected to use frequencies in the 10 to 30 MHz range, which has been used for decades by amateur radio operators, as well as international shortwave broadcasters and a variety of communications systems (military, aeronautical, etc.). Power lines are unshielded and will act as antennas for the signals they carry, and have the potential to completely wipe out the usefulness of the 10 to 30 MHz range for shortwave communications purposes.

Modern BPL systems use OFDM modulation which allows the mitigation of interference with radio services by removing specific frequencies used. A 2001 joint study by the ARRL and HomePlug powerline alliance showed that modems using this technique “in general that with moderate separation of the antenna from the structure containing the HomePlug signal that interference was barely perceptible” and interference only happened when the “antenna was physically close to the power lines”.

Much higher speed transmissions using microwave frequencies transmitted via a newly discovered surface wave propagation mechanism called E-Line have been demonstrated using only a single power line conductor. These systems have shown the potential for symmetric and full duplex communication well in excess of 1 Gbit/s in each direction. Multiple WiFi channels with simultaneous analog television in the 2.4 and 5.3 GHz unlicensed bands have been demonstrated operating over a single medium voltage line. Furthermore, because it can operate anywhere in the 100 MHz – 10 GHz region, this technology can completely avoid the interference issues associated with utilizing shared spectrum while offering the greater flexibility for modulation and protocols found for any other type of microwave system.

And What About Hams (Amateur Radio Operators) opposing BPL? 

(It’s important to understand that Hams oppose BPL interference to wireless spectrum, and do not oppose BPL itself or broadband, despite what several BPL carriers and vendors may have said publicly.)Q: Why are Amateur Radio operators, also known as Hams, in an uproar over Broadband Over Powerline or BPL?
A: BPL is a system that is being tested to provide broadband Internet service via powerlines. The system uses radio frequencies that do radiate into the air. They can cause interference to licensed services including Amateur Radio. This interference has been seen and documented in most of the BPL trial areas. BPL experimental licenses have been issued allowing some systems to operate between 2 and 80 MHz (megahertz). This particular band of frequencies are generally known as HF (which is actually 3 – 30 Mhz) or “shortwave” frequencies. This part of the radio spectrum has very special properties not found elsewhere. With this band, one can communicate around the world with very low power levels, without the need for any equipment in between, such as satellites or repeaters. This is due to the fact that radio waves in this band can bounce off the ionosphere multiple times to get to the destination. Other portions of the radio spectrum are essentially line-of-sight. This means that the signals cannot bend or bounce off the ionosphere, but they can only propagate like light – in a straight line. Many consumers have enjoyed listening to shortwave radio broadcasts from around the world.

Q: What is technically wrong with BPL?
A: The medium of BPL (the powerline cable), unlike any other broadband medium (copper twisted pair, fiber, coaxial cable), is inherently unsuited for carrying the frequencies BPL uses. Power lines, copper twisted pair, and coaxial cable all act like natural low pass filters, meaning higher frequencies are attenuated more than lower frequencies when attempting to transmit them through the medium. The exact slope of the graph of attenuation depends on the specific construction of the material, but in general, twisted pair is suitable up to 100 Mhz and coaxial cable can go up to about 3 Ghz. Again, these are very general figures and determining the suitability for any application depends on other factors. Power lines would be suitable for up to perhaps 20 Khz, maybe 350 kHz at a stretch, with caveats. The exact figure is unimportant for this discussion, but note that this is kilohertz, not megahertz or gigahertz. These are essentially audio frequencies, and equate to a data rate in the neighborhood of ISDN.

Power lines are designed to carry electrical power. They were not designed to carry radio signals. They do this very poorly, loosing much of the signal to losses and, more importantly, radiating them as radio signals that can and do affect nearby receivers using those frequencies. Amateur radio operators, CB operators and shortwave listeners are all found commonly in the residential neighborhoods where BPL will be installed. They will all suffer strong interference if BPL uses their frequencies at the permitted levels. Other uses of HF spectrum include business, government, military and aeronautical. Many of these users and their organizations have expressed strong concern about BPL and its interference potential.

Other wires and cable, such as telephone or cable TV systems do actually radiate to some extent, but in proportion to the amplitude of the signal they are carrying, it is minuscule. (This doesn’t apply to poorly designed or maintained cable or pairs, or intentionally radiating cable like Radiax used for indoor applications). It’s ironic that many antennas used on HF are very close in construction and visual appearance to typical power lines. It has been clearly demonstrated in technical studies prepared by ARRL, the National Organization for Amateur Radio that BPL often causes interference and that power lines act like antennas.

Putting a signal that will operate 24 hours a day over wide range of frequencies at the FCC limits for unlicensed devices onto wiring that radiates well is a very poor engineering choice. When that system is built as large as an entire community – or several states –noise and interference are inevitable. Basic first-year college courses teach one this. This is very, very basic. The electrically dirty and unpredictable nature of power lines can also produce harmonics and intermodulation that can cause unexpected forms of interference.

Q: Does BPL work?
A: From a consumer point of view and what has been heard from test areas, yes. Amateur radio operators and others do not have an issue with this. So far, the BPL systems installed are generally small, and how well they will work when built out to cover a wider area and more customers remains to be seen, though, as there are few if any heavily loaded systems. However, one would expect the behavior of a BPL network to resemble that of a shared medium like cable or wireless. Systems will have to be segmented further as traffic grows and the available bandwidth in a segment decreases.

There are significant concerns about immunity to interference which are discussed later in this FAQ. Tests done to date show that nearby radio transmitters can interfere with BPL, but this too, needs to be looked at in larger systems. The effects of such interference are unknown at this point and could make BPL service in areas unreliable or unfeasible. Under the FCC rules on BPL, as an unlicensed device, BPL is offered no protection from licensed users that may interfere with BPL in the normal course of their licensed operation.

Q: What are the benefits of BPL?
A: The dream of BPL is to have every powerline activated with BPL and have BPL networks within homes. With the advent of inexpensive chipsets, every appliance in the home could easily be networked without additional wiring. Utilities could use BPL to manage network elements within the power grid and perform automatic meter reading or AMR, although narrowband non-interfering PLC based AMR system existed previous to BPL and continue to be in use today. Grid management functions could be performed as well, although PLC and SCADA systems do this today.

One major perceived advantage of BPL is wiring infrastructure. Utility power is nearly everywhere, so costly copper twisted pair or fiber would not have to be run. While this may be seen as a major advantage, the work and cost to light up every power line with BPL is significant. Furthermore, nearly every home in the U.S. has twisted pair copper lines installed for telephone service. These lines are better suited for broadband transmission and are the basis for DSL, so technically the potential coverage and benefits of DSL far outweight that of BPL and the power line infrastructure.

Q: Is BPL the same as wireless broadband?
A: No. BPL affects wireless radio spectrum but it does not actually use the wireless spectrum through the air to transmit data. There are some BPL systems that use unlicensed 802.11 WiFi wireless as part of the BPL system to deliver the data to the end user, however the core BPL network uses the powerline as the base medium for transmission.

Q: Why can’t BPL coexist with wireless technologies?
A: To allow a wired-based network to make large portions of frequencies unusable in the RF world is tantamount to the way industry used to be permitted to cause serious pollution. Cable companies use frequencies in VHF and UHF bands that were they to leak out into the outside world, would wreak havoc with aviation and public safety frequencies. They are subject to strict limits that include a need for them to regularly test the cumulative effect of the leakage from their system and there’s no reason why any other wired network like BPL shouldn’t be subject to strict limits – it’s a wired medium, it is not necessary for it to affect wireless media in any way.

Q: Is BPL new?
A: BPL has been tested and deployed on a limited basis in other countries. BPL vendors may claim “new technology” and advances have now made it possible, but the fact is they can’t change the laws of physics. High speed data must occupy a certain amount of “bandwidth” and power lines which were designed to operate at 60Hz will radiate RF that is applied to them. The BPL system has to operate at a higher level than the noise on the lines, so by definition, BPL will increase the radiated noise level from the power lines that carry it, on any spectrum it used. Only changing power line construction (i.e. coaxial cable) would eliminate this radiation. BPL proponents reject this as being too costly, but that would be the cost to make this a real viable technology.

Q: Hasn’t Power Line Carrier or PLC been in operation for years without problems?
A: The “original” PLC is a very low frequency, narrow bandwidth signal used for control equipment in the power grid. It is in the neighborhood of 100 to 180 kHz (that’s kilohertz, not megahertz). It is not intended for high speed data transfer, but rather simple commands, like “turn relay on”, “turn relay off”. It is also used to send “messages” up the line when a major failure occurs so that other network elements don’t trip off when senses a fault. This keeps the power grid from falling down like a bunch of dominoes (like what happen recently with the blackout in the Northeast :-).

Usually people refer to the 1-80 Mhz “PLC” as BPL, although some use the acronyms interchangeably and in my opinion, wrongly. Comparing “original” PLC and BPL is like comparing apples to oranges. PLC is brief, primitive commands and is very narrow banded. BPL is wide band noise that’s all over the place.

Recently the Amateur Radio community petitioned the FCC for a low frequency band in the same area as the original low frequency telemetry PLC. It was rejected by the FCC because utility companies complained that Amateurs would interfere with this telemetry PLC. This is very interesting for several reasons. First off, the utilities are saying that BPL won’t interfere with Amateurs and Amateurs won’t interfere with BPL. But PLC is arguably more robust than BPL because it’s narrow band. So which way is it ? The second reason this was so interesting (or disturbing) is that telemetry PLC is operating under Part 15 and is afforded no protection from interference from licensed services. For the first time the FCC disallowed licensed operation and essentially protected a Part 15 operation. This sets a dangerous precedent as any unlicensed “spectrum squatter” can later claim rights to a chunk of frequencies. The third and even worse observation is that the utilities have admitted that the telemetry system of the national power grid is vulnerable to attack, and they’ve based a critical part of the infrastructure on something that cannot be legally protected from interference. Any nutcase with a little knowledge could conceivably control network elements within a power grid with enough time and money.

PLC has been in operation for a long time, but as mentioned, it’s very different from BPL. It seems many BPL proponents are confusing the situation by riding on PLC’s history and merits.

Q: Haven’t power lines always radiated radio energy and caused interference?
A: Previous to BPL, power lines have had a history of radiating noise, but properly maintained lines will radiate only a minute amount of radio energy that is low enough that it does not impact radio communications. On the other hand, a properly maintained BPL system will radiate radio energy.

A distinction between power transmission noise and BPL noise needs to made, and is very important to understanding this issue. Power transmission noise is a product of 60 hertz voltage and is a rough, raspy frying type of noise that is stronger at low frequencies and gradually tapers off. Two types of BPL noise exist. One is results from Spread Spectrum modulation which is a Geiger counter type noise. The other type of BPL noise typically observed is resulting from OFDM modulation. This creates carriers, or little “radio blips” every 1 kilohertz or so across the radio spectrum. Each “blip” has a ringing sound or clicking sound.

Power transmission noise can most always be fixed by changing out worn or defective power line components such as insulators or taps. Radio interference from BPL is a side effect of a the system in operation, not an anomaly.

Ironically, BPL proponents and FCC representatives have claimed that BPL deployments have reduced classic power line transmission noise. This is true to an extent as any noise on the line from power transmission will seriously degrade a BPL signal. Unfortunately, the power line transmission noise is being replaced with a much more devastating signal that is stronger and does not decrease in strength as you go up in frequency.

Q: Does replacing bad or dirty insulators on the powerline fix the BPL interference problem?
A: Replacing such components reduces power transmission generated noise, but not the BPL interference. Often new BPL carriers have to clean up their lines and reduce the power transmission generated noise so BPL will work. BPL is a radio signal on the powerline, and basic electronics theory states that any unshielded conductor a ¼ wavelength long will radiate RF energy.

Q: BPL seems to be more prevalent in Europe. Is BPL interference just a problem in the United States? Is the interference issue related to North American 60 hertz systems versus 50 hertz systems found in the rest of the world?
A: The BPL interference problem is a function of the unshielded powerline, not the frequency of the power signals, 50 or 60 hertz. European systems (50 hz) tend to be better suited for BPL than North American systems (60 hz) from a logistical point of view as utilities in Europe tend to place tens, or a hundred or more customers on one transformer. In the US there is usually only two, three, or four customers on one transformer. Because European voltage service to the home is 240V versus North American 120V service, the I2R losses in transformers and drops in Europe are one quarter that of North American systems, thus more homes can be placed on one transformer. More customers on one transformer translates to less transformer bypasses that are needed to pass the BPL signals to household wiring.

Again, the feasibility of BPL service is not related directly to power service frequency.

Q: Won’t adaptive technology in BPL protect others from interference?
A: Adaptive technology was proposed by the FCC in the Notice of Proposed Rulemaking in 2003 as a way to mitigate interference. While sounding high tech, this technology is actually administrative functions that are present in most BPL equipment today. This includes:

• Power control, dynamic or remote
• Frequency notching
• Harmful interference shut-down feature

First off, these techniques do not provide any protection for mobile or portable stations. Dynamic power control is an obvious no-brainer and should be a requirement as this would keep power levels on the lines as low as possible. Frequency notching, while a possible solution for local interference complaints, doesn’t address long range interference that would be created by ionospheric propagation or the cumulative effect on the noise floor in the HF bands. The harmful interference shut-down feature is rather ambiguous in the NPRM, but it seems to be a manual remote control on/off switch. Most network equipment today can be turned off remotely, so this feature isn’t a real stretch, but it does nothing proactively to lessen interference. Also, many uses of shortwave involve more listening than transmitting, and uses like international shortwave broadcast would not have any transmitters nearby for the BPL system to sense.

Q: If this just affects Ham Radio, why should anyone care?
A: BPL will also affect other licensed services such as government agencies, military, aviation, maritime, public safety, and shortwave broadcasts. Ham radio occupies less than 10 percent of the affected radio spectrum. Many of these users have filed comments in the FCC rulemaking, in both the Notice of Inquiry and the Notice of Proposed Rulemaking, that expresses grave concern about BPL interference.

Q: Are Hams qualified to talk about BPL issues?
A: Hams are licensed by the FCC in the United States and various governing bodies in most every country in the world. Most countries have several classes of licenses. In the United States and most countries, one must pass written tests on electronic theory, communications protocol, and regulatory material. While becoming a licensed Ham isn’t equivalent to an Electrical Engineering degree, many hams have formal educations in electrical engineering.

The experience and knowledge of Hams can vary greatly, and like any hobby or profession, there are hams who speak before thinking. Occasionally you will find an ex-ham dismissing Ham Radio as being dead, or an inexperienced Ham who only uses VHF bands who could care less about the HF bands that are threatened by BPL. Most active hams though are rather knowledgeable. Their decades of experience on radio give them an understanding about interference that is unmatched in any arena.

Q: Why are Hams the only people talking about the negative affects of BPL? Why aren’t other services complaining?
A: Ham radio occupies a minority of BPL spectrum, with government, shortwave, public safety, and ship communications occupying the majority. Ham radio operators are the most vocal because government agency employees quite simply don’t post in Internet forums. Many of these other users are concerned about BPL interference. Comments with the FCC have been filed from groups and agencies like the NTIA, Salvation Army, the Missouri State Patrol, Aeronautical Radio, Inc, and NPR, to name just a few.

Most of the other services on HF radio frequencies affected by BPL like government and aviation usually operate at much higher power, use much larger/more directional antenna arrays, are more frequency agile, and often have their transmit/receive stations at sea, in the air, or on remote reservations of land far away from civilian housing areas and their associated power lines. In the latter case, having acres and acres of open land gives much more flexibility about antenna placement than someone living on a 1/4 acre residential lot. Full realization of each of these advantages is not attainable by even the richest radio amateur. So even if government agencies such as the NTIA say that BPL will be of minimal impact to them, it should be emphasized that their capabilities, and therefore their operating environment, may be very different than that of amateurs.

Q: How can HF radio signals travel across the world?
A: Long distance HF propagation occurs by radio signals bouncing between the earth and the ionosphere, often several times.
This has two consequences with regards to BPL. A lot of long distance wireless HF communications occurs just above the noise floor (the common noise you hear on your radio when you are not receiving a station which is noise leftover from the Big Bang and from various man made sources), receiving very weak signals. A station attempting to receive an HF signal could experience interference from a local BPL system, even if it is notched in the amateur radio bands and 40 or 50 dB below the emissions limit. (A good analogy is it’s hard to hear someone yelling a half mile away when someone is wispering in your ears). The second issue has yet to be proven either way, but with thousands or millions of BPL devices in operation, the noise from these devices combine could raise the noise floor in the HF spectrum and propagate thousands of miles. The jury is still out on this and the NTIA was supposed to release a second study that would explore this. There’s not nearly enough BPL systems in operation to test this in real world circumstances.

Q: Isn’t long distance HF communications more an infrequent anomaly rather than a common occurrence?
A: HF propagation in an area varies on the time of day and on solar activity. It’s usually possible to communicate on some HF band to various places in the world 24 hours a day. During the day, the D, E, and F layers form a thicker layer of ionization. This ticker layer absorbs lower frequencies (below 5 Mhz or so), and enables farther propagation of frequencies between 15 and 30 Mhz. At night, the D, E, and F layers combine. This causes the upper frequecy limit (called Maximum Usable Frequency or MUF) to drop, usually to about 14 or 15 Mhz. During this time, frequencies below 5 Mhz will propagate better. This is why you can hear many AM broadcast radio stations at night, and most AM radio stations decrease their power at night to avoid interference.

Q: Why doesn’t my 802.11 WiFi or cellular phone radio signal travel around the world?
A: 802.11 WiFi uses 2.4 Ghz frequencies (2400 Mhz) which is considered microwave frequencies, cellular is 800 Mhz, and PCS is around 1.3 Ghz. These to not bounce off of the ionosphere, but travel right through it. The reflective characteristics of the ionosphere diminish above about 30 Mhz. Also, microwave frequencies are much more susceptible to absorption by precipitation and water vapor.

Q: Does BPL Affect Shortwave Listeners (SWLs) and Shortwave Broadcast Frequencies?
A: Yes, BPL can affect all of the HF shortwave broadcast bands.

Q: Can’t Hams and others using the spectrum simply be relocated?
A: Users of the affected spectrum cannot be relocated, or at least not economically or in a timely manner. It would be cheaper for the government to subsidize cable and DSL deployment. Plus, all of the services that use HF bands require the characteristics that only HF bands exhibit. There would also be huge international treaty implications with any relocation. Changes in international communications treaties are measured in decades, not months or even years. Relocating government and military services alone would take years as the FCC would have to structure a migration plan. Chances are it would be ten years before this could be completed and it’s likely that power companies will have run fiber to the home or DSL and cable will finally be ubiquitous. Perhaps the largest issue to tackle, though, is where to move these services in what is an already overcrowded spectrum.

If it was determined that relocation was the way to go, this would be very irresponsible as HF radio bands are a unique natural resource. No other radio spectrum can provide worldwide communications without any supporting infrastructure (i.e. satellites).

Q: Why doesn’t the FCC allocate dedicated spectrum for BPL?
A: There are two main reasons for this. The first is quite simply there is not enough open space in the HF spectrum to accommodate a dedicated allocation for BPL. Current BPL systems need at least ten MHz of spectrum. Newer, faster systems may need 25 MHz or more to operate. Accommodating such systems would require the entire HF spectrum to be allocated to BPL. The second reason against a dedicated frequency allocation is that BPL doesn’t actually use the wireless spectrum, it emits energy in the frequency range, thus polluting it. An analogy would be building a highway and dedicating it to one industry or group of people so they could dump garbage on it.

Q: Won’t Frequency Notching protect licensed services?
A: Frequency Notching is a new feature that is reported on some of the second generation BPL systems. The idea is that if a BPL signal is interfering, the system can be configured to not use this frequency or a band of frequencies. This apparently can be done on a subscriber, network, or system wide basis. While this is a nice feature, it still doesn’t make BPL acceptable for four reasons. The first of which is that the possibility of intermodulation still exists. (Intermodulation is described in detail below).

The second reason is that BPL signals can propagate for very long distances due to the characteristics of the frequencies they are using. Interference from a BPL could be experienced a thousand miles away. Tracking this interference down so that the BPL operator could be informed of a Part 15 violation in order to get them to notch the frequency would be logistically difficult. Considering BPL acts like a large distributed antenna and not a classic point-source of interference, it would be hard to direction find the signal to track it down.

The third issue is what’s called the noise floor. The noise floor in simple terms is what you hear in between radio stations on an FM or AM radio. It’s the snowy screen you see on your TV when there’s no station on the channel. The noise floor is essentially noise and radio spectrum energy left over from the Big Bang. This noise is fairly constant, but man-made noise sources such as existing 60 Hz power lines, noisy transmitters and other things contribute to raising the noise floor. BPL systems will contribute to the noise floor and raise it higher. This will make it more difficult to receive weak signals. Taken to the extreme, the raised noise floor could make HF communications impossible.

The fourth reason that notching won’t work is that there just isn’t any significant open space in the HF spectrum. Anywhere you look, there’s a service using the frequency. My guess is BPL providers are banking on the fact that there’s not a local user of particular chunks of HF frequencies, so they will configure their systems to operate in these areas. If we were talking about microwave frequencies that are strictly line-of-sight propagation bands this would be a great solution, but that’s simply not the case with HF. HF has worldwide propagation characteristics, so it’s likely users hundreds or thousands of miles away could experience interference. As mentioned elsewhere here, tracking such wide band no identification interference is nearly impossible.

Q: What is intermodulation?
A: Intermodulation is the mixing of radio signals which produces new radio signals. Think of it as radio waves having children. But just how do radio waves have children ?

This mixing is caused by what are called non-linearities. One non-linear electronic component that you find in most any electronic device is a diode. When multiple radio signals are run through the diode, they mix together. Let’s say we have a 4 Mhz signal and a 6 Mhz signal going into the diode. We would then get:
4 + 6 = 10 Mhz
6 – 4 = 2 Mhz
4 and 6 Mhz had two “children”, 2 and 10 Mhz !

Now, non-linearities are usually good. This phenomenon is used in just about every radio device to either create a signal to be transmitted, or receive a signal that you hear or see. But, non-linearities can occur where you don’t want them and then in causes problems. One such place is in power lines. Bad, corroded connections or dissimilar metals touching can create natural diodes that act like mixers and produce this intermodulation.

So, let’s take a BPL signal and for the sake of discussion, say it’s a grossly simplified consisting of radio signals at 1, 5, 8, 9, and 12 Mhz. Some of the intermodulation products that could be created would be:

1 + 5 = 6 Mhz
8 + 9 = 17 Mhz
9 + 12 = 21 Mhz
12 – 9 = 3 Mhz

But you could also have what is known as third order products:

1 + 9 + 12 = 22 Mhz
8 + 9 + 12 = 29 Mhz
8 – 5 + 12 = 15 Mhz

Or even:

2 * 12 = 24 Mhz
(9 – 5) * 12 = 48 Mhz

You can do the math and figure out each permutation, but you get the idea. If we took a real BPL signal that has signals from 1 – 80 Mhz the number of products and where they would fall are mind-boggling. The resulting intermodulation products in a system could extend well above the band BPL proponents want, falling into FM broadcast, VHF TV, Aeronautical, and more public safety bands. This is just another reason why BPL is so problematic.

It’s arguable that such non-linearities in power lines are exhibited as arcing connections, something that most power companies are actively searching for these days as the RFI (radio frequency interference) effects are well understood. These maintenance issues will be addressed quickly by well run utilities. However, non-linear loads are common in homes, light dimmers being the first devices that come to mind. Theoretically, these devices could create intermodulation that would in turn be radiated by the house wiring and outside power cabling.

Q: What are harmonics?
A: Harmonics are in the same “family” as intermodulation. It’s new radio signals that are created by a non-linear device that are a multiple of a radio signal. Unlike intermodulation which requires two or more “parent” or fundamental signals to be created, harmonics are the product of just one signal. A 10 megahertz signal would have a second harmonic of 20 Mhz, a third harmonic of 30 Mhz, a fourth harmonic of 40 Mhz, and so on. Harmonics created in equipment are generally undesirable and can interfere with other communications services if not filtered out. Harmonics can also mix with other signals to form intermodulation.
Harmonics can be created from fundamental BPL signals on powerlines. These have been observed in the field at typically 30 dB lower than primary BPL signals.

(compiled using informations from Wikipedia and http://bplinterference.wikispaces.com/)

• 1 Comment

PPC ads like Google AdSense and Clicksor are a great source of income for many web site publishers.

You sign up and put the ads on your web sites. They deliver the ads live, with each page load — ads relevant to the content of the page the ads are delivered to. (Well, sometimes they aren’t so relevant, but they try.)

The ads are delivered from the inventory of ads they have available.

What happens when no ads are available depends on the PPC ad distributor. AdSense delivers non-profit (non-paying) ads. I don’t know what Clicksor does.

Both of those organizations, however, allow you to specify content to display whenever they don’t have sufficient relevant ads in their inventory. Instead of their ads, they display your ads.

I will show step by step how to make those alternate ads so they fit within the space allotted for the PPC ads.

You’ll be able to deliver your own ads, paying ads or your personal selection of non-profit ads, whenever AdSense or Clicksor inventory runs out.

Step 1
Create a web page that will contain your ad. A style in the head area should specify zero margins.

Here is an example:

<html>
<head>
<style type=”text/css”>
body { margin:0px; font-family: sans-serif; }
p { font-size:14px; line-height:18px; }
a:link { text-decoration: none; }
a:visited { text-decoration: none; }
a:hover { text-decoration: none; }
a:active { text-decoration: none; }
</style>
</head>
<body bgcolor=”white”>
<p>
This is the best!
</p>
<p>
<a href=”http://example.com”>Click Here!</a>
</p>
</body>
</html>

The ad on the web page can contain any content a normal web page could contain — text, images, Java, forms, and so forth — so long as it fits within the allotted ad space.

I’ll show you an easy way to tweak for the fit.

Step 2

Make a separate, temporary web page to be used for tweaking the ad web page. The temporary web page will contain an iframe tag, which is the “secret” for quickly and easily adjusting the content on the ad web page.

The iframe tag is the same dimension as the allotted ad space. Let’s assume a 250 x 250 square space for our example:

<html>
<body>
<iframe
src=”http://example.com/adpage.html”
width=”250″
height=”250″
frameborder=”1″
marginheight=”0″
marginwidth=”0″
scrolling=”no”>
</iframe>
</body>
</html>

The iframe has a border so it’s easier to tell how much more space the ad web page can occupy, or where the excess content is cut off.

Put your ad page URL in the src attribute (in place of http://example.com/adpage.html).

Adjust the width and height when designing ads of different sizes.

(This temporary web page may be kept on your computer’s hard drive, handy for when you want to tweak the size of another web page ad.)

Step 3

Load the temporary web page into your browser. The ad web page will load into the iframe.

Adjust the ad web page as needed until it fits nicely in the iframe. Test in various browsers, especially the two most popular ones, IE and Firefox. Different browsers render content a bit differently, even when CSS styles are used.

Step 4

Upload the ad web page to your server and make a note of its URL.

Step 5

Tell the PPC ad distributor the URL of your alternate ad web page. This is done by specifying the URL in the JavaScript code that delivers the ad.

The following apply as of this writing.

AdSense:

For AdSense ads, the ad web page URL can be specified when you generate the ad code. The URL will then appear in the generated JavaScript between the quote characters of this line:

google_alternate_ad_url = "";

Clicksor:

For Clicksor ads, put the URL of the ad web page between the single-quote characters of this line:

clicksor_default_url = '';

If the line isn’t in the code obtained from Clicksor, it can be inserted.

Ads of Different Sizes

Make an alternate ad web page for each PPC ad size on your web site.

Once set up, you can display ads that make you money even when the distributor is out of inventory.

• 0 Comments

We hear a lot about “Web 2.0″ these days. It sounds neat and it’s trendy to talk about blogging and social media. But does it really affect our businesses? Is Web 2.0 just for kids and tech-hipsters or is it something we business owners should use to help promote our businesses?

I can’t tell you if Web 2.0 is right for your business but I can tell you it’s something to be aware of Ignoring it means ignoring a possible tool that could be valuable in helping you get more customers.

So, to help you get started in thinking about Web 2.0 for your business, here are some things for you to consider.

• 0 Comments

“Data Breaches Are Frequent, but Evidence of Resulting Identity Theft Is Limited; However, the Full Extent Is Unknown…” This is what United States Government Accountability (GAO) says in a report to Congressional Requesters.

Why GAO Did This Study?
As a result of advances in computer technology and electronic storage, in recent years, many entities in the private, public, and government sectors have reported the loss or theft of sensitive personal information. These breaches have raised concerns in part because they can result in identity theft—either account fraud (such as misuse of credit card numbers) or unauthorized creation of new accounts (such as opening a credit card in someone else’s name). Many states have enacted laws requiring entities that experience breaches to notify affected individuals, and Congress is considering legislation that would establish a national breach notification requirement. Policymakers, consumer advocates, and others have raised concerns that data breaches can contribute to identity theft, in which an individual’s sensitive personal information is used fraudulently. The Federal Trade Commission (FTC), which is responsible for taking complaints from victims and sharing them with law enforcement agencies, has noted that identity theft is a serious problem—millions of Americans are affected each year, and victims may face substantial costs and time to repair the damage to their good name and credit record.
Although there is no commonly agreed-upon definition, the term “data breach” generally refers to an organization’s unauthorized or unintentional exposure, disclosure, or loss of sensitive personal information, which can include personally identifiable information such as Social Security numbers (SSN) or financial information such as credit card numbers. Data breaches can take many forms and do not necessarily lead to identity theft. The term “identity theft” is broad and encompasses many types of criminal activities, including fraud on existing accounts—such as unauthorized use of a stolen credit card number—or fraudulent creation of new accounts—such as using stolen data to open a credit card account in someone else’s name. Depending on the type of information compromised and how it is misused, identity theft victims can face a range of potential harm, from the inconvenience of having a credit card reissued to substantial financial losses and damaged credit ratings.

Beginning with California in 2002, at least 36 states have enacted breach notification laws—that is, laws that require certain entities that experience a data breach to notify individuals whose personal information was lost or stolen. There is no federal statute that requires most companies or other entities to notify affected individuals of data breaches, although federal banking regulatory agencies have issued guidance on breach notification to the banks, thrifts, and credit unions they supervise. In addition, the Office of Management and Budget has issued guidance—developed by the President’s Identity Theft Task Force—on responding to data breaches at federal agencies. Because a number of bills have been introduced in Congress that would establish a national breach notification requirement, Congress asked GAO to review the costs and benefits of such a requirement and the link between data breaches and identity theft.

GAO was asked to examine (1) the incidence and circumstances of breaches of sensitive personal information; (2) the extent to which such breaches have resulted in identity theft; and (3) the potential benefits, costs, and challenges associated with breach notification requirements. To address these objectives, GAO reviewed available reports on data breaches, analyzed 24 large data breaches, and gathered information from federal and state government agencies, researchers, consumer advocates, and others.

This report focuses on breaches of sensitive personal data that can be used to commit identity theft, and not on breaches of other sensitive data, such as medical records or proprietary business information. To address the first two objectives, GAO obtained and analyzed information on data breaches that have been reported in the media and aggregated by three private research and advocacy organizations, as well as information on breaches collected by state agencies in New York and North Carolina, federal banking regulators, and federal law enforcement agencies. GAO also collected information on breaches experienced by federal agencies compiled by the House Government Reform Committee in 2006 and by the Department of Homeland Security (DHS). In addition, GAO conducted a literature search of relevant articles, reports, and studies. GAO also conducted interviews with, and obtained documents from, representatives of federal agencies, including the FTC, the Department of Justice, DHS, and the federal banking regulatory agencies; selected state government agencies and the National Association of Attorneys General; private and nonprofit research organizations; and consumer protection and privacy advocacy groups. Further, we obtained information from industry and trade associations representing key sectors—including financial services, retail sales, higher education, health care, and information services—that have experienced data breaches. In addition, for the second objective, we examined the 24 largest (in terms of number of records breached) data breaches reported by the news media from January 2000 through June 2005 and tracked by private groups. For each of these breaches, GAO reviewed media reports and other publicly available information, and conducted interviews, where possible, with representatives of the entities that experienced the breaches, in an attempt to identify any known instances of identity theft that resulted from the breaches. GAO also examined five breaches that involved federal agencies, which were selected because they represented a variety of different circumstances. For the third objective, GAO reviewed the federal banking regulatory agencies’ proposed and final guidance related to breach notification, and interviewed representatives of each agency regarding their consideration of potential costs, benefits, and challenges during development of the guidance. Further, GAO reviewed the strategic plan and other documents issued by the President’s Identity Theft Task Force. In addition, GAO conducted a review of the effects of California’s breach notification law, which included interviewing and gathering information from California state officials and selected California companies, educational institutions, and other entities subject to the law’s notification requirements.

Read the Full Report Here

• 0 Comments

SQL Plan Cache Snapshot, Visual Explain, and other tools solve a DB2 for i5/OS performance mystery.
By Jim Flanagan and Shantan Kethireddy

The story in this column illustrates a methodology DBAs can use to solve SQL query performance issues in DB2 for i5/OS V5R4 when crucial evidence is missing.

Mr. D.B. Admin of Company A wants to hire famed detective Shantan Kolmes and his sidekick Dr. Flatson to help with a serious performance problem. As of last week, Company A’s main order application has been experiencing drastic performance degradation. Mr. D.B. Admin calls Kolmes and Flatson to explain the problem.

“Every day at 10 a.m., our customers’ orders are at their peak, and our operators can’t process them fast enough. We’ve had customers disconnect mid-order. We’ve lost several long-time customers and who knows how many new customers. Our chief competitor, Company Z, is quickly taking over several of our accounts. Performance wasn’t an issue until very recently. Please help — the CEO is ready to fire the entire IT staff!”

“Do you have a SQL Performance Monitor from when the system was running well?” asks Kolmes.

“No, we didn’t see the need for it.”

“I suppose you don’t have an SQL Plan Cache Snapshot either?” Kolmes says.

“No, does that matter?”

“It will make things a bit more difficult.” But Kolmes remained confident. “Has anyone changed the application since you started seeing this problem?” Kolmes asks.

“No, all changes are supposed to be approved through me.”

Kolmes finds Mr. Admin’s choice of words curious. “You say, ‘supposed to.’ It sounds like you might have had the same issue in the past?”

“Well, yes. I have one developer who is a bit of a maverick. But he assures me he didn’t change anything.”

“Have you noticed anything else unusual?” Kolmes asks.

“We did have a contract programmer recently for a minor update, but I approved that change myself.”

Kolmes says he’ll arrive before 10 a.m. the next day. “Whatever you do, please don’t IPL the system!”

In the meantime, Flatson researches Company Z. As he looks through his notes from the questioning of Mr. Admin, Kolmes begins to develop a preliminary strategy for clue collection and analysis.
Mr. Kolmes and Dr. Flatson arrive at Company A at 9 the next morning. Mr. Admin reports that he successfully stopped the system operator from performing an IPL of the system.

Snapshot Significance

If you’re in the situation facing Mr. Kolmes, the first order of business is to capture the SQL performance environment by taking a SQL Plan Cache Snapshot. To do so, go to the iSeries Navigator interface, expand the Databases folder and the specific database folder, and then right-click it. Select the SQL Plan Cache option and then the Show Statements menu action.

On the statements screen, you can fill in the filters you’re interested in. In the Kolmes case, we know which application is causing the problems, so you’d fill in the schema name for the queries that use or reference these objects. You can also set the filter for Top ‘n’ queries with the largest total accumulated runtime, if you were sure you’d catch the queries for KEYPROD1. Figure 1 shows what Kolmes retrieved. You can save the list of queries to a schema for later viewing by choosing Create Snapshot and giving the schema and snapshot name.

Figure 1. List of queries referencing KEYPROD1.

Figure 1 shows the names of the contract programmer, Mr. Admin, and the maverick programmer. But row six shows something interesting; the User Name in that row matches the last name of Company Z’s CIO.

Notice how easily we (and the two detectives) uncovered preliminary evidence. Before V5R4, you had to turn on a monitor during a period when the application was experiencing difficulties and then analyze the monitor data using monitor analysis queries. This approach could cause system performance degradation if appropriate filtering wasn’t used. You also had to try to duplicate the environment in which the application was running. Now, you can simply take a SQL Plan Cache Snapshot with no system overhead, and have all of the same information and analysis tools at your disposal. Keep in mind that the SQL Plan Cache only contains information about queries that went down the SQL Query Engine (SQE). In-depth analysis of queries that went down the Classic Query Engine (CQE) may still require use of the Database Monitor. See Resources for more information about the difference between SQE and CQE.

If we had had a SQL Plan Cache Snapshot or a monitor from before the point of the initial degradation, we could have run the comparison menu action between the old and new environment to see what was changed. Instead, we’ll need to analyze the worst running queries from the application under the present conditions, figure out the root cause of the performance bottleneck, and then try to figure out what changed (and why) to cause this bottleneck.

A Visual Clue

In order to view the detailed performance information for a given query, you would use a tool called Visual Explain. Visual Explain is a database tool that graphically represents the optimizer implementation of a query request. It provides a method of identifying and analyzing database performance problems. The tree-like output provides a comprehensive understanding of the optimized plan. Visual Explain shows all the attributes necessary for users and developers to understand implementations, performance bottlenecks, indexing strategies, and so on.

Going back to the SQL Plan Cache viewer dialog, you can simply select the desired statement and click Run Visual Explain.

The Final Select icon shows all environmental attributes such as pool size, number of processors, QAQQINI information, and so on. Based on the list of environmental settings Mr. Admin provided, the detectives find these attributes to be in order. Mr. Admin’s company doesn’t have SMP installed, so the lack of parallel processing is expected. You can see the area of the tree that was estimated to be the most costly by going to the View menu and selecting the Highlight Index Advised menu option. In Mr. Admin’s case, the most costly area of the tree was a table scan on the two-billion-row ORDERS table, which populated a hash table used for joining.

The Index Advisor (available as an icon in the toolbar) suggests indexes to improve performance. In Mr. Admin’s case, the Index Advisor suggests an index over a heavily used join column from the two-billion-row ORDERS table, which would explain the poor performance.

If you look at the rest of Mr. Admin’s poorly performing statements, you see that all the statements reference the same join column from the same ORDERS table. To find out if this missing index is behind the performance problems, you have to determine whether this index, or a comparable index, had existed before the performance degradation and that it was removed immediately prior to the degradation.

Reconstructing Details

To find a record of objects in the database from a period when the application was running well, you would open the iSeries Navigator application and expand the Systems folder. Right click on the database folder and select the Health Center menu option. You should see the Overview page. Click on the Size Limits tab. In Mr. Admin’s system, the settings are already set to the schema the application is running in. Hit Refresh to show the current size limits for KEYPROD1. If you want to save the results on this screen, choose save, and the save dialog displays with the history file name filled in.

Now that the current environment is saved to the history file, you can do a comparison to the last capture. Click View History… and fill in the history filename.

When Mr. Kolmes sees the history file that contains the two collections of the Size Limits for this application, he expands the Maximum number of index entries node in the two collections so that he can compare the history files of the current and last months (see Figure 2).

Figure 2. Comparison of current and past history files.

His examination shows that there is one less index in the current environment than in the previous environment. The missing index is IMPORTANT_IDX1.”

Kolmes tells Flatson that they need one final piece of data: the auditing journal (QSYS/QAUDJRN). Using the Display Journal (DSPJRN) command on the auditing journal reveals the precise moment when the index was removed — and who removed it.

Who Done It?

Mr. Kolmes calls a meeting with Company A’s CEO, Mr. Admin, the system operator, and the maverick programmer.

“We found a DROP INDEX entry in the audit journal. Does anyone know why that was done?”

The system operator says it sounds just like the maverick programmer.

Kolmes replies, “We thought that might be the case, but the queries in the SQL Plan Cache Snapshot show your User Name, Mr. System Operator.”

“That’s not unusual; I do queries when diagnosing a system problem.”

“Ah, but your name also shows up in the audit journal as the User Name who did the DROP INDEX. We also found out that the CIO of Company Z has a nephew that works in this company,” Flatson chimes in.

The system operator yells out, “It was me alright, I admit it! I would have gotten away with it too if it hadn’t been for those new tools, and if you hadn’t stopped me from IPLing the system.”

“Tell it to the judge,” Kolmes replies.

---

In Mr. Admin’s case, a company wanted to steal a competitor’s business, so they used an insider to remove a key index. Once the index is added back, order processing performance should return to its expected sub-second timeframe.

The moral of the story? Keep an SQL Plan Cache Snapshot of your environment so you have something to check against if you encounter performance problems in the future.

Jim Flanagan is an advisory software engineer at IBM in database development for DB2 for i5/OS and team leader of iSeries Navigator — Database.

Shantan Kethireddy is a member of the IBM SQL Query Engine team and a frequent speaker at System i technical conferences.

• 0 Comments