Sunday, September 27, 2009

Where Did All Those Gorgeous Russians Come From?

Though this is a fairly frivolous question (OK, extremely frivolous), I am convinced it has an interesting answer. To put it bluntly, in the Soviet Union there was no market for female beauty. No fashion magazines featured beautiful women, since there weren't any fashion magazines. No TV series depended upon beautiful women for high ratings, since there weren't any ratings. There weren't many men rich enough to seek out beautiful women and marry them, and foreign men couldn't get the right sort of visa. There were a few film stars, of course, but some of the most famous—I'm thinking of Lyubov Orlova, alleged to be Stalin's favorite actress—were wholesome and cheerful rather than sultry and stunning. Unusual beauty, like unusual genius, was considered highly suspicious in the Soviet Union and its satellite people's republics.

Ultimately, what goes for the fashion world goes for other spheres of human activity. In the past, you had to play chess or be a champion gymnast to come to international attention if you were born in the Eastern bloc—chess and competitive sports figuring among the few party-approved export industries. Nowadays, stars in fields previously unsanctioned by the party—crime novelists, conceptual artists, computer whizzes—from Russia, Hungary, or Uzbekistan have a shot at fame and fortune, too. As for talented entrepreneurs, the sky's the limit.

Beauty is a matter of luck, but the same could be said of many other talents. And what open markets do for beautiful women they also do for other sorts of genius. So, cheer up next time you see a Siberian blonde dominating male attention at the far end of the table: The same mechanisms that brought her to your dinner party might one day bring you the Ukrainian doctor who cures your cancer or the Polish stockbroker who makes your fortune.

Saturday, September 26, 2009

How Do You Create Config Files Automatically?

"When deploying new server/servergroup/cluster to your IT infrastructure, deployment (simplified) consist of following steps: OS installation: to do it over network, boot server must be configured for this new server/servergroup/cluster; configuration/package management: configuration server has to be aware of the newcomer(s); monitoring and alerting: monitoring software must be reconfigured; and performance metrics: a tool for collecting data must be reconfigured. There are many excellent software solutions for those particular jobs, say configuration management (Puppet, Chef, cfengine, bcfg2), monitoring hosts and services (Nagios, Zabbix, OpenNMS, Zenoss, etc) and performance metrics (Ganglia, etc.). But each of these tools has to be configured independently or at least configuration has to be generated. What tools do you use to achieve this? For example, when you have to deploy a new server, how do you create configs for, let's say, PXE boot server, Puppet, Nagios and Ganglia, at once?"

Actually this is one of the goals VMWare is proposing to meet with their vSphere. vCenter, ad nauseum initiatives. [full disclosure I've beta'ed VMWare software since v1]. This also presupposes full P2V, V2P cross machine conversions if required. The goal here is be anywhere, and run anywhere.

Now if I had the money, I'd toss full de-dup into the storage array mix as well, so much of the image file size essentially disappears unless there is simply no duplication anywhere. And if you are in that situation, take my advice. Quit, or just shoot yourself and get it over with.

It's been a long time since I played at that level (six mainframes, eighteen mini's, 575 desktops, and I never got an accurate count of the 100+ laptops) but at some point you have to ask yourself, when does the customization end? Standardization was the only thing that kept myself and my team of four !relatively! sane.

If you seriously need customization of that level, then you aren't doing things right. Reduce each VM to a single app (Apache, MySQL, IIS, network appliance, whatever) and use virtual switches to create a topology as required. Think of each VM as a particular Lego block, or IC: Systems Componentization as it were. And this is where de-dupe will also shine.

Friday, September 25, 2009

What Have VCs Really Done for Innovation?

.... So I’m miffed by the National Venture Capital Association’s (NVCA) claim that companies like Microsoft and Google “…would not exist today without the funding and guidance provided during their early stages by venture capitalists.” And I’m amused that the NVCA claims credit for creating 12 million jobs and generating $3 trillion in revenue (that’s only 21 percent of U.S. GDP). In the software industry (which includes Internet/Web 2.0), they stake claim to 81% of the all jobs created. Yes, 81%. Can they please give the entrepreneurs who risk their life savings, max out their credit cards and put their families in the back seat a little more credit? We’re not talking about divvying up the company’s stock here, just a pat on the back.

How’d they come up with these numbers? They added up all the revenue generated in 2008 by any company a venture capitalist ever invested a dime in. So if John Doerr bought Bill a lunch in 1985, they’d count Microsoft as part of their empire. Maybe I’m exaggerating a bit. But seriously, the NVCA numbers aren’t even remotely credible. How can VCs claim credit for the revenue of a company which they cashed out of twenty or thirty years ago? And even then, claiming credit for 81% of tech jobs and 21% of GDP? More to the point, would those jobs never have been created if the VCs had never appeared on the scene? How can the NVCA prove causality?

The answer is, the NVCA can prove nothing and a growing pool of data suggests that VCs at best have little to no impact on these companies and at worst have a negative impact. I just completed a research project in which we interviewed the founders of 549 successful companies in several high-growth industries – the ones VC’s are most likely to fund. We selected companies that had made it out of the garage and were generating real revenue. Guess what? Hardly ten percent of the serial entrepreneurs took venture money in their first startups. In their subsequent launches, the proportion who took venture money went up to a quarter. In other words, three-quarters of even the most experienced entrepreneurs didn’t rely on venture capital (new report to be released in October).

Kids Who Get Spanked

The IQs of the younger children who were spanked were 5 points lower on average four years later than those of children of the same age who were not spanked. Scores among the older children were an average of 2.8 points lower among spanked children than children who were not spanked.

"Parents spank to decrease bad behavior in the short and long term and to promote positive behavior," she tells WebMD. "What the research tells us is that spanking doesn't seem to be doing either of these things."

Friday, September 18, 2009


The fundamental problem here isn't the RAID concept, is that the throughput and access times of spinning rust haven't changed much in 30 years. Fundamentally, today's hard drive is no more than 100 times as fast (both in throughput and latency) than a 1980s one, while it holds well over 1 million times more.

ZFS (and other advanced filesystems) will now do partial reconstruction of a failed drive (that is, they don't have to bit copy the entire drive, only the parts which are used), which helps. But there are still problems. ZFS's pathological case results in rebuild times of 2-3 WEEKS for a 1TB drive in a RAID-Z (similar to RAID-5). It's all due to the horribly small throughput, maximum IOPs, and latency of the hard drive.

SSDs, on the other hand, are no where near the problem. They've got considerably more throughput than a hard drive, and, more importantly, THOUSANDS of times better IOPS. Frankly, more than any other reason, I expect the significant IOPS of the SSD to signal the death knell of HDs in the next decade. By 2020, expect HDs to be gone from everything, even in places where HDs still have better GB/$. The rebuild rates and maintenance of HDs simply can't compete with flash.

Note: IOPS = I/O Per Second, or the number of read/write operations (irregardless of size) which a disk can service. HDs top out around 350, consumer SSDs do under 10,000, and high-end SSDs can do up to 100,000.

Thursday, September 17, 2009

Stock Market Manipulation By Millisecond Trading

A firm I worked with recently tore down an arbitrage network (they were getting out of the business as it was not core) which comprised of a great deal of Layer 2 dark fiber between sites in NYC and an external data center in NJ, Force 10 fabric switches with multiple paths to server clusters, and a great many Sun X-series servers running Linux. This arbitrage network bypassed the standard corporate (i.e. Cisco-based) network as they wanted exclusivity, higher bandwidth and as much speed as possible. Still, there were issues and the whole environment was scrapped since the actual returns did not match the expectations or cover the costs.

When I looked over the shoulders of the designers (they didn't want too much support from the regular network engineering team) they were concerned with raw performance and not as much with security or other daily operational issues. I would characterize it as the difference between, say, a NASCAR Sprint Cup car and your regular transportation. The former is purpose-built solely for performance while the other has to contend with safety requirements, daily functionality, and a lower common denominator for use.

The location of their server no longer matters as GS was allowed to put a peeker in line between everyone else (on the planet) and the publicly traded ETNS. They get an opportunity to front run every transaction. Every single one. Zerohedge is a financial blog that has been keeping up with this story. Their coverage is good if a little breathless.

The Securities and Exchange Commission is seeking to curb a practice criticized for giving an unfair advantage to some market participants who have lightning-fast computer trading software.

Friday, September 11, 2009

I’m Better Than You Are

... she thanked me profusely for the rescue and also for not saying “That’s almost as bad as…” When I asked her what she meant, she asked me had I never noticed how people always say they’ve had worse things happen to them than you have? This was the first time I’d met someone else who was bothered by this.

I have occasionally noticed this pattern in speech before and while it’s sometimes humorous, more often than not it's annoying. Only after she reminded me, though, did I stop to analyze why we do it.

I can understand when you are telling a person you had an unexpected expense that was tough to deal with and they reply that they can relate, then go on to tell you of their most recent unexpected outlay. That, I think, is their way of saying they have been through what you’re going through. Another phrase is, of course, “Been there!”

But what I’m really talking about is the one-upmanship or ‘I’m better than you are’ meaning behind that phrase "almost as bad."

"Oh yeah, that’s almost as bad as…”

Now, there may be times when indeed, my experience might not have been as bad as one they’ve been through, but even in that case, isn’t it still wrong to begin dialog with someone in that way?


.. the systematic and conscious practice of "creative intimidation", making one's associates feel inferior and thereby gaining the status of being "one-up" on them.

Wednesday, September 02, 2009

Costco exceptional cash conversion cycle

Costco also has an exceptional cash conversion cycle. That's a little formula that measures, in days, how quickly a company can buy inventory, get it on the shelves, and sell it, thus converting it to cash. Costco turns over its entire inventory 12.8 times a year (more than once a month!) -- allowing it to sell merchandise even before it has to pay its suppliers for it.

This enables the company to buy some of its inventory on the vendors' payment terms instead of using its working capital. Costco's outstanding cash conversion cycle has run one to three days over the past few years. By comparison, a more traditional retailer like Macy's usually takes at least 70 days to convert its inventory to cash.