Slashdot Mirror

That was (is) my system until I had the flu and the outage was at 3am. I didn't think I was too capable of dragging the gen outside in the rain and doing the extension cords... I just powered down the systems.

Some UPS manufacturers like APC spec a max load, but allow for adding additional "battery-only" units. What's not entirely clear to me has been whether or not I could use deep-cycle marine batteries (alot cheaper), plugged-in to one of their controllers....Just a thought though.

In all battery and generator backups, you will find cheap bargains but the output is "sloppy" -- not the sine-wave most devices expect. Some devices won't care, but more sensitive electronics that have average or sub-standard power supplies might.

Honda advertises their EU inverter line as being clean for electronics use -- clean in the sense of pure 50 or 60Hz sine wave output. Not a square, triangle or rhomboid shaped output.

The SmartUPS 1400 model pictured is a 24 volt system, so 2 serially connected deep-cycle lead acid batteries must be used.

ups1
ups2

This setup worked great during the recent hurricanes and power outages in Florida. It kept my laptop, cell phones and flashlights charged during the outages, the longest of which was about 5 days. During prolonged outages you would want to turn the UPS off, and only run it for an hour a day or so to charge smaller devices, check the news, etc.

It is extremely quiet (only audible humming comes from two small fans on the case) which is important to me. It is also very low on energy consumption. I got an APC Back-UPS ES-350 (just a couple of days before the big black-out here, in North-East USA --- could not have been wiser :) The UPS is rated at 8 minutes under 100W load and 2 minutes under 200W but it lasts over 40 minutes powering my server and my DSL modem.

Another thing I am really happy about is the fact that VIA seems to be doing a good job supporting Linux. Personally, I have never had trouble running Red Hat on mine (although, I hear FC2 had issues with it that were only recently fixed --- but that was FC2's problem).

Overall, I feel that this has been a really great product and would wholeheartedly recommend it. I am also very happy to see that VIA has been constantly improving them. I am looking forward to seeing the upcoming nano-ITX boards.

Take a look at APC's rackmount "ProtectNet" stuff.

A 1U rack mount chassis with 24 slots (you can protect up to 16 data lines) is $30. Then you can buy different plug-in modules for different devices. They have them for 10/100BaseT, regular Telco phone lines, T1/ISDN/etc, RS232, etc.

Get one of these for $18 per Cat5 you want to protect.

Keep in mind that nothing is going to protect against a direct lightning strike, but these are good filters for surges that can come from an indirect hit.

Take a look at APC's rackmount "ProtectNet" stuff.

A 1U rack mount chassis with 24 slots (you can protect up to 16 data lines) is $30. Then you can buy different plug-in modules for different devices. They have them for 10/100BaseT, regular Telco phone lines, T1/ISDN/etc, RS232, etc.

Get one of these for $18 per Cat5 you want to protect.

Keep in mind that nothing is going to protect against a direct lightning strike, but these are good filters for surges that can come from an indirect hit.

clicky

My ups ran my fridge for three days in the summer, there is no reason i can think of that it cant handle my computer for a week.

My ups ran my fridge for three days in the summer, there is no reason i can think of that it cant handle my computer for a week.

This is not so much of a problem if you're not completely forsaken braindead.

You need lightning protectors like this (if the link doesn't come up, search for "Blitzschutz", and of course, ground the access point. If lightning hits, the antenna is toast, as well as the protectors, but the rest of the network is OK. What I suggest as well as to protect is the ethernet plug, there are protectors for this as well, like this, but I'm sure there are others as well.

I just found out that the 12 blackouts a week that I was having was because the little powerbox, that joins the wire from the power pole to my house, had shit itself. Diagnosis confirmed because my house was the only one powerless, and my mains switch and all fuses were still on (up). They (ETSA) were quite sure that the power box had died of old age, and that the plank that it was screwed to and that the wires ran through was all wet and rotten at the back had nothing to do with its untimely demise. The workers all seemed very cheerful though it was late on Sunday afternoon and raining. Maybe the double overtime rate had something to do with it?

Since deregulation of our power supply (Adelaide South Australia), blackouts have become a regular occurance, especially during really hot weather or windy weather. Can't we make airconditioners that run using heat, the same way as a kero fridge does? And obviously there isn't enough money to fund maintenance of the wires or pruning of trees. So everything is falling apart.

Not to mention, that this house has about one power point per bedroom, and just two in the office. In 30 years since this house was built, I've gone from one powerpoint in the bedroom for a reading light, to one each for the light, the stereo, the fan heater, the phone charger the AA battery charger, alarm clock, mozzie zapper, hair dryer etc. And don't get me started on the room full of computers. Etsa/AGL are charging double (instead of less as they promised) and I'm using triple.

What the APC guy wrote:

>I would suggest a new product we have available which is the BR1000I. This UPS is sufficient to support two PC's, two monitors and the 8 port hub. Because laser printers draw so much power, the laser printer will need to be plugged into the Surge Only outlet at the rear of the UPS, as it will not be able to be held up on battery power. >

>Recommended Retail Pricing for the BR1000I is $919.00. Below is a link to the spec page of the BR500I.[wadiwood: how did we get from BR1000I to BR500I?] >

> BR1000I

Maybe I should just get a diesel generator. Or imagine the sleek athletic bod I'd have if I hooked the computer to pedal power?

It doesn't. This does.

(Simple Solution - APC Power StruXure)

First off, get rid of the stand-alone UPS systems. They are inefficient and not reliable from a system standpoint. You are much better off with a single 30kW unit rather than ten or twenty 1,000VA units. With single UPS, it is easy to either install a local panelboard (US), or a busbar distribution system (UK).

Second, anything that you do to enhance flexibility will compromise reliability. If that is an acceptable risk they go ahead and do something modular.

Given these constraints, a 30kW UPS is 100A or so three phase, 208V or 150A single-phase 208V. 120V would not be appropriate; 480V is a cheaper input feeder, more hassle for bypass.

The best solution might be to go the traditional panelboard route, and run branch circuits to all your equipment in liquidtight flexible non-metalic conduit (the blue stuff).

Alternately, you can install a 100A plug-in busbar the length of your cage and tap off 30A, 208V, 1PH for each rack. They are designed for light industrial applications, but work pretty well if you can sacrifice on reliability.

The APC solution is neat because you don't need an electrician, and it is all modular. Might not be as flexible as the component route, but it works. I think it includes managed outlets in the racks as well.

Get a 220v APC UPS with a PDU. Run a fewBaytech RPC2s out of the APC PDU.

Doing this myself actually.

hmm, I have several of the office models myself (BF500U, etc.). accroding to apc's site, they're not included in the recall:

recall faq

This right here is where you'll find all the details about the safety recall, and what you need to do to arrange for a replacement.

No you think before posting...
The CS models are USB only connectivity models.
And the apcupsd people have figured out how to talk to them. And you can read the model number and serial number and a bunch of other stuff off of them.

And I happen to have one of the affected units.
[root@stream RPMS]# apcaccess | egrep 'APCMODEL|SERIALNO'
SERIALNO : AB0142147520
APCMODEL : Back-UPS 500

As usual slashdot provides the worst URL for the story. The URL in the story is simply the press release. This is their main site about it, has much better information about how to identify if your UPS is part of the recall... and links to a nice FAQ. Of course I'm still wondering what type of unit they will replace mine with.

I've deployed a lot of small data centers. Your choices in what to buy are actually easier than you think. You can calculate your power and cooling needs down to the last BTU and KVa, and then you have to buy equipment in a size they make them in! Just go the smaller end of the lines.

You'll end up with a 3 ton or 5 ton air conditioner. Liebert air conditioners can also humidify/dehumidify and heat/cool. There is a market for used Lieberts if you want to save some money. Call your local A/C contractor.

A 3Kva UPS would be a good size unless you want more standby time in which case you could go for a 5 Kva or maybe two 3's for redundancy.

Liebert makes great UPSs, too. The APC Matrix line is a pretty good design because you can hot-swap the batteries yourself.

To cool your mini server room try an APC Network Air system, they have really huge expensive ones, and then this one: apc.com. Little expensive, but it cools the room down and takes moisture out of the air. They have a vent duct you have to run outside though, and you'll want to get the kit to hook the dehumidifier to a pipe or something that drains outside, because at .26Gallons of water per hour, with its tiny reservoir it can fill up fast, although it seems like it shuts off the dehumidifier if the tank gets full.

-Ryan

Aspects of this page indicate it's not yet released. For instance, lots of stuff is XX'd out; and if you click on "Fuel Cells" in the nav bar, you get a notice implying that the product is not yet ready.

Is it possible that this is not the final pricing? It could be an early number, could be the very top (so nobody claims "false advertising" if they stumble across it later, when they set the real price), could be misinformation for competitors, whatever.

Oh, nobody's mentioned numbers yet, but to get a single data point, you can get an APC's Matrix 3000XR (which sustains 500kW for about 5:15, and is in many ways more capable-- higher peak, for instance-- but obviously-- can't be refueled during a power outage). It's listed at $3750 US.

Eww! Tripp-Lite?

Throw those in the garbage where they belong (but recycle the batteries!) and install something decent...

Eww! Tripp-Lite?

Throw those in the garbage where they belong (but recycle the batteries!) and install something decent...

If you go to American Power Conversion's website and look around, they have some good information on why running your servers at 230 is more efficient than 115 anyway.

Quick and simple

Quick and simple.

On another topic: Your comment violated the postercomment compression filter. Comment aborted

Lose the lameness filter, Taco.

What good would internet access be without power, dude?

Plenty of good.

Anyway I had a lot of flakeyness when I first got it. Turns out a lot of the problems were with getting the BIOS settings right. I had to slow the AGP port down to 2X, disable the onboard IDE controller (not in use), disable some of the power conservation features, and tweaked a half dozen other settings. Now the system is solid as a rock, even when I boot into Windows. (Win2000Pro, not ME/98) Haven't crashed anything in probably 6 months.

I've been building my own machines for about 10 years now. It's always the same, getting the right combination of BIOS settings, system settings, and stable software is an iterative process that just takes time. My machines are usually unstable for the first 2-4 months after I get them until I can stomp out all the bugs. Once in a while you run into a real dud piece of hardware and have to bite the bullet on it, but usually some persistance pays off. It does take time though.

I do take certain precautions now though to ensure a stable and easy to maintain system, not all of which are cheap but all help a lot. Some of these include

• I always use SCSI. Yeah IDE is cheaper and it works just fine, but if you can afford it, SCSI is better in pretty much every aspect except price. Easier to set up, easier to expand, the drivers are usually better, it's faster (at the high end anyway), etc. If price isn't a huge issue, get SCSI.
  
• I purchase a UPS. (I like APC personally.) They are cheap insurance. They will protect your machine from brief power dips which are often more damaging than spikes, they carry an insurance policy on them if you equipment is damaged, and they unless you want auto shutdowns, they don't have any compatibility issues.
  
• I get a good power supply. PC Power and Cooling is my brand of choice. This is slightly paranoid but I've had two machine have power supplies go south on them. When the power supply goes, it usually fries things. So I get one that is very unlikely to go.
  
• Don't get no name components if you can avoid it, and preferably get something you've worked with before. I use Adaptec controllers because I worked with them for many years, they work and I know what to expect. Not to say others don't work great but I know what I'm getting and that makes the inevitable debugging easier. It's also more likely that drivers will be available if more than 3 people have that piece of equipment.
  
• I avoid equipment with unusual interfaces. If you have SCSI, get as much SCSI equipment as is practical. Don't mix and match interfaces if you can avoid it.
  
• I generally avoid integrated controllers or disable them if I get one. Onboard SCSI and ATAPI can work fine but are usually harder to work with and troubleshoot. Driver availability tends to be not as good as a rule though this is not always true.
  
• I don't overclock. Not that there is anything wrong with doing it, but don't expect a stable system if you overclock. You might get it but chances are fairly high you won't too.
  

APC has something like a $25,000 guarantee for power surges. I personally have seen an APC surge protector melted to a puddle of goo, and the equipment plugged into it was just fine after a new power cord. (The end melted into the surge protector.) I am absolutely sure that the $50 more you spend on a good surge protector is worth every penny. APC also makes UPS units with the same capability (as far as surge protection), as well as other line protectors such as CAT5, phone, and COAX. Protect every line going into your datacenter and if anthing gets hit by lightning, APC will write you a check to repair the equipment. See their site for the details.

I wouldn't put any X10 stuff on my production systems, I have NOC monkeys that can push reset buttons for me.. That stuff is purely for fun... and they are fun to play with, may not be secure, but neither is my telephone or television, someone can walk by my house and broadcast propaganda on my TV(if I didn't have cable.;).

If I wanted a serious, secure, remote power control system, I'd go out and buy one, I wouldn't trust anything I built myself.

APC makes a good one, SNMP able and ready. I've used this on remote test stations and they work well. Expensive though.

most network printers are all direct-IP addressable and have little need for a print server

Don't discount the humble printer server yet. The fairly expensive IP-connected printer my company has, uses a dodgy implementation of TCP such that when it is receiving a print job, it stops listening on the lpd port. Printing a document then becomes a hit-and-miss affair.

We once installed a HP hardware print sharer box for a customer. Unreliable as hell, and probably for the same reasons.

I suspect that other embedded-TCP devices use similar "cut-down" implementations. The APC masterswitch is vulnerable to ping-o-death and probably other attacks, such as synfloods. TCP is best left to "real hosts", IMHO.

ON THE HIGH END?? You must be kidding. Out of the narrow selection of AMD motherboards, they all have stability problems, and their performance over Pentiums is negligible.

Cites, please? Do you have anything to back up this claim?

If you are having stability problems with an Athalon system, it's probably because you failed to follow AMD's guidelines. The Athalon, particuarly early ones (like mine), are finicky about the hardware they work with. They are particuarly sensitive to the power supply, which is why AMD has a list of recommended power supplies on their web site. A UPS with a power conditioner is a smart investment for any high-end system. (I use an APC Smart-UPS 650)According to the guys I buy my hardware from, somthing like 90% of the stability problems with Athalon systems that they see come from people using an out-of-spec power supply. Most of the rest of the problems come from using marginal memory.

My primary system is a FIC SD11 with an Athalon 550. I got it within the first two weeks of it hitting the market. It's almost a year old now, and I have had ZERO stability problems with it. It runs 24x7; the only time I ever have to reboot it is when I switch into 95 to do some gaming. This box primarily ran NT4 Workstation up until May, when I switched it over to RedHat 6.1 It was rock-solid even under NT, and has been just as stable under Linux.

It makes NO sense whatsoever to buy a top-of-the-line CPU and MoBo and stick it in a bargain-basement case with a cheesy power supply and no-name RAM. Spend the extra money and get server-grade memory and power. Likewise, if you ignore the manufacturer's guidelines and use out-of-spec parts, you have no right to be pissed at them when your substandard components crash the system.

Before you build an Athalon system, do yourself a favor and RTFM first. You'll save yourself a lot of aggrivation.

"The axiom 'An honest man has nothing to fear from the police'

No slam at you, spazmoid (I hate the decision, too), but the only "of course" here is the fact that we are moving into a genuinely new realm, and there are going to be some dislocations, as we sort things out.

In the past, you could have two identical brand names (for example) in nearby town, state, or even adjacent aisles of the grocery store (would you even notice if there was an 'Encore' brand dog food, and an 'Encore' line of TV dinners [feel free to provide you own punch lines.

In the past, boundaries were much more effective at compartmentlizing business, and the primary role of the trademark was associative -- i.e. it relied on the associational mechanisms of the brain more than the actual literal words.

On the web, we rely on literals, and any hierarchy of naming will fall short of the flexibility of the old system (any system that had more flexibility than ordinary human associative thinking would be useless because humans wouldn't be able to master it) We do not live hierarchically, we do not do business hierarchically (i.e. category-busting businesses and products will always emerge), and we do not want to be ruled by a hierarchy.

Meanwhile, the fact that there is *one web* is both a strength and a weakness. It's strengths are well known, and many of its weaknesses lie in the fact that it will require us to change our thinking. This is often a necessary price for new capabilities, but it's always a drawback from the user's view, in the short term (and we're *all* users, when it comes to the web, whether third grader or multinational)

I don't much care for this decision, and I sure wouldn't enjoy debating whether the Corinthian Sailing Club (corinthian.org) is more 'worthy' of that second level domain than the guy who lost corinthian.com.

In the future, we will see new ways of doing things, and one that I expect will become more common will be "index pages" that list, for example, many 'corinthians' and link their websites. We've had examples of this for years:
until recently http://www.apc.com let you choose between the former American Power Corporation, a well-known UPS maker, or the former Alabama Power Corporation (an electric utility and management company)

My general plan is to ignore that there are humans anywhere around. They usually do the wrong thing and so they've been taught to keep there hands off the entire thing.

I never use more than 50% of the UPSes rated capacity. It gets better run times out of them and stresses them less. Often the UPSes are the only remaining power at a place and the temperatures in these places are often into the 90s while the UPSes are still doing their thing. Less load means they can effectively withstand more heat before they reach the end of the batteries. Further, things that matter more are more apt to be on their own UPSes, so they last the longest (given similar sized UPSes).

I avoid APC UPSes, since several I've had them fail in odd and unexpected ways (in particular boiling batteries so they reach end-of-life way too early.) APC is hitting a price point, not a quality point and has been unfriendly to the OSS movement. I've had very good luck with Best and SOLA

If you run Unix (or FreeBSD or Linux or...) I suggest UPSD to baby sit your systems. Most of the sites I take care of are effectively 'lights out' (i.e. nobody that is a systems person is there regularly) and UPSD has served me well. Power fails at site X and UPSD emails me. I can then call (being up to a continent away) and manage the problem as need be. Don't forget to UPS the phone system.

I've found that most of the time the outages come in two possible groups. 0-30 minutes and 2.5hrs and more. I make sure that all of the 0-30 kind of events are fully protected. Without generators the 2.5 hr kinds are not practically covered by normal UPSes. If you're prone to lots of power failures, then go look at Home Power Magazine and in particular Trace Engineering systems. The theme is generate your own power 24/7 and be much more reliable as long as one does some minimal maintenance.

Finally, all the UPSes are tested annually for run times and the batteries are replaced at slightly longer intervals than the manuals suggest. Also be sure the that users have flashlights or other emergency lights that work. I test those lights too, since no one else gives a damn about emergency lighting.