Home ยป consolidation


Central Hardware, St. Louis history

St. Louis-based Central Hardware was one of the first big-box home improvement chains. It peaked in 1993 at 39 stores in six states in the midwest, employing 3,700 people. It was once the 19th largest hardware retailer in the United States.

Central Hardware’s motto was “everything from scoop to nuts,” a play on the English idiom “soup to nuts,” which means beginning to end. Their inventory was over 40,000 SKUs, comparable to today’s home improvement stores. Its stores regularly exceeded 50,000 square feet. That’s about half the size of a typical home improvement store today, but it was large for the 1970s and 1980s. Traditional hardware stores ranged in size from 2,000 to 10,000 square feet. Its employees wore orange vests so customers knew who to ask for help.

Read More »Central Hardware, St. Louis history

90s computer brands

Some 90s computer brands are the same as today, but a lot more companies played in the field than now. Profit margins were higher then, so industry consolidation wasn’t the matter of survival that it is now.

Here’s a look back at some of the brands of old, including some famous PC brands, some not-so-famous, and some notorious. The 1990s were certainly a make or break time for many of them.

Read More »90s computer brands

First impressions of VMWare

I’ve been setting up VMWare ESX Server at work, and it’s quirky, but I like it. I shut it down improperly once (logging into the console on its Linux-based host OS and doing a shutdown -h now resulted in a system that wouldn’t boot anymore) so I’m afraid of what may happen. The upside is since every virtual machine is just a collection of files, disaster recovery is dirt simple: Build a VMWare box, restore those files from backup, point the VMs at them, and you’re back in business. No more need to worry about locating identical or close-enough-to-identical hardware. For that reason alone, I’d advocate running all of my Windows servers in production environments on VMWare, since Windows isn’t like a real OS that will allow you use a disk or image on dissimilar hardware with minor adjustments. We get some other benefits too, like allowing us to put all the toy servers in one box with RAID to protect us in a disk crash. We’ve lost far too much to disk failures on desktop PCs recast as someone’s pet-project server.
It also appears to try to only allocate to machines the amount of memory they’re actually using, so theoretically, if you were doing server consolidation and had, say, four servers with 256 MB of RAM, you could potentially get away with putting them on a VMWare host with less than 1 GB of memory.

I also like VMWare for tasks you don’t like to dedicate a single machine to. For instance, DNS on NT is totally brain-dead. It’s slow to propogate. It works about 99.9% of the time, but that .1% of the time that it feeds wrong answers will infuriate somebody, who will holler at you, and the struggle to fix the problem will infuriate you worse.

If you want DNS that works, your best bet is to load Linux or BSD with BIND on something and use it. But if you don’t already have a production Linux server somewhere and you don’t have a machine you trust to give the job, carve out a server on a VMWare box. Allocate 16 megs of RAM and a couple hundred megs of disk space to it, and give it a thin slice of processor time. DNS lookups don’t take a lot of power, so it won’t detract noticeably from the other hosted servers.

It ain’t cheap (the price isn’t listed on the web site for a reason), but software’s cheaper than hardware.

I’m back.

Very interesting. Just as everyone’s proclaiming Linux dead, Red Hat goes and turns a profit for the first time. Yes, there are too many Linux companies. Yes, there’ll be consolidation. No, I’m not convinced that selling it at retail is necessarily the best way to proliferate the system.
I also find it humorous that people like ZDNet’s David Coursey can struggle all weekend setting up a Windows server, yet state that Linux is no threat to Microsoft, even as a server. The implication is that Linux is too difficult. Give me a weekend–actually, more like 5 minutes, if you’ll spot me TurboLinux and a 50X CD-ROM drive–and I can have DNS going on Linux, easy. Give me a day, and I can have a lovely mail server going too. (I intended to do that just this past weekend, actually, but I couldn’t come up with a working ISA SCSI controller to pair up with my army of SCSI CD-ROMs to make it happen.)

Needless to say, this past week I lost most of what little respect I had for Coursey. VMWare runs Windows under Linux better than VirtualPC runs Windows on the Mac, and Coursey’s obviously never heard of it (see that second link).

Don’t get me wrong, Linux setups drive me up the wall sometimes. But I’ve had instances where Windows flat out wouldn’t install on perfectly good hardware, for no good reason, too. And since Linux servers are unencumbered by a GUI, multimedia, Pinball, Internet Exploiter, and other desktop stupidity that has no business on servers, they’re a whole lot easier to troubleshoot. You’ve got a kernel, a daemon or two, and a plaintext configuration file. That’s not much to break. Actually it’s good engineering–a machine should have no unnecessary parts.

So long, Cal Ripken. Cal Ripken announced he’s hanging it up yesterday morning. I had the pleasure of seeing Ripken play shortstop a couple of times in the early 1990s when the Orioles were in Kansas City. Today, in this era of A-Rod and Nomar and Jeter, Ripken’s offensive stats don’t seem so hot. But in the 1980s (and before), if your shortstop could hit .270 and steal the occasional base, you counted yourself very, very lucky. In those days, Ripken not only hit .270, he was consistently one of the best defensive shortstops in the American League. He was never as flashy as Ozzie Smith, but how many shortstops ever fielded .996? You’re happy to get that kind of a fielding percentage out of your first baseman, and first base is the easiest position to play. Not only that, Ripken was also good for 20-25 homers and 80+ RBIs. These days that doesn’t sound too impressive either, but remember that Ripken played the bulk of his career in an era when people rarely hit 40 homers–someone who could pop 30 was considered a real power threat.

And besides all that, Ripken played 2,632 consecutive games, shattering Lou Gehrig’s record of 2,130. Ripken played the majority of those games at shortstop (he also played some third base at the beginning and at the end). Gehrig played his games at first base and in left field, both much less demanding positions. And while Gehrig played every inning of every game just once, Ripken did it four times, in consecutive years (1983-1986).

Ripken’s really slowed down the past three years, but he did end his streak on his own terms before being cut down by injuries his final three seasons. He’s nowhere near the player he used to be. Then again, at the end of his career, Ernie Banks couldn’t hit or field, and he was playing first base. Ripken refuses to move from third to a less demanding position–partly out of pride, but partly because he’s still capable of playing third.

And we can’t forget his loyalty. Ripken’s played his entire career, from 1981 up until now, with Baltimore. You don’t see that much anymore.

Hardware developments

Hardware news. Lots of stuff today. We’ll take it one at a time.
AMD to hit 1.5 GHz by January. Intel intends to release a 1.5 GHz P4 in late November/early December. AMD’s a bit behind that (assuming Intel will deliver, which they’ve been having difficulty doing lately), but the 1 GHz Athlon performs similarly to a 1.4 GHz P4. Good news for us, bad news for Intel. AMD intends to release a 1.2 GHz Athlon within a month, along with an 800 MHz Duron.

In related news, the AMD 760 chipset (the SMP-enabled one) will be released this year.

The P4 problems are related to the use of PCI graphics cards and Intel has reportedly fixed the problem. Although allowing much higher clock rates, the P4 is less efficient than the P3, so a 1.4 GHz P4 is expected to give comparable performance to a 1 GHz P3. It won’t be until Intel hits 1.5 GHz and higher that the new architecture will give any performance advantage over what’s available now. Not that you can find a 1 GHz P3…

Memory prices are down. If you’re looking to buy, this is a good time. You never know when they’ll rise again, or fall for that matter.

Maxtor buys Quantum. In a consolidation of disk manufacturers, Maxtor bought the disk manufacturing wing of Quantum for $2 billion, making Maxtor the world’s biggest disk manufacturer. Quantum’s tape operations will be spun off into a new company, to be named Quantum.