Computers & Electronics

UPS backup power supply worth it?

  • Last Updated:
  • Jul 22nd, 2019 5:48 pm
Member
Jul 6, 2009
384 posts
110 upvotes
warpdrive wrote: If the computer is writing to critical system files and you shut it off (due to a power loss), you may not be able to recover your system easily. At best case, the software may detect that the operation failed and recover itself.
Cited is a fear eliminated over 20 years ago. That problem observed by many who still kept using FAT filesystems (ie Win 95, Win ME). That technology had been obsoleted before OSes existed. The fewer and informed back then never used those "obsolete when first introduced" OSes. Instead used Windows NT.

If power is lost due to a power outage, then "not obsolete" file systems ignore an incomplete file save and restore a previous file. System recovery is automatic.

Worse are so many technically naive who insist other already saved data can be corrupted by a power off. It is not. It was not corrupted by a power off even when disk drives moved heads with motor oil. Power off does not destroy saved data. Power off is only a threat to unsaved data.

One condition that could corrupt one file during a write was eliminate by technology that was standard over a generation ago. And still that myth lives on.

UPS provides temporary and 'dirty' power so that unsaved data can be saved. And to avert an inconvenient reboot.
Nothing more.
Deal Expert
User avatar
Mar 6, 2003
19712 posts
13576 upvotes
Ottawa
What you said is false. Modern journaling file systems in current OSs will do a lot ensure filesystem integrity but it doesn't prevent all instances of data corruption. This distinction is important because it's beyond just losing remaining unsaved data. Differing disk caching mechanisms can also affect the integrity of data if there is a power loss.

Personally I have observed perfectly functioning computers on our site fail to easily recover after a power outage. It's far beyond an inconvenient reboot.

I'll take my real world practical experience over your ramblings thank you.
Please update your profile to include your city https://forums.redflagdeals.com/ucp.php?i=ucp_profile&mode=profile_info
Member
Dec 7, 2015
479 posts
105 upvotes
Ottawa, ON
Gee wrote: Check the battery.
Unit reports the battery is fine and everything is hunky-dory. It's only a year old, so I wouldn't expect a problem. There have been lots of power hiccups over the year, but only one extended outage and it shut everything down well before the battery was drained - nothing to suggest the battery has been beat to death.

One interesting thing was that the clock in my kitchen, which always goes to a blinking 00:00 state whenever there's a power glitch, ran through this just fine. It was more of a brownout (actually two quick flickers of the lights) than either a normal brownout or a complete shutoff. It may be a condition the UPS just didn't want to handle.
Member
Jul 6, 2009
384 posts
110 upvotes
warpdrive wrote: Computers running SSDs can be even scarier when power failures occur.
That is a problem with a design of some SSDs. Whereas that would not cause a problem to (most) disk drives in the 1960s and 70s, that same unacceptable fault has been noted in some SSD designs.

Return for a minute back to the 1970s. Some hardware was designed badly. For example, if the eight inch floppy disk was in the drive when power cycled, then that floppy could be erased. This was directly traceable to how 5 and 12 volts were controlled. Yes, some hardware can be that badly designed. But that is not the fault of a power loss or brownout. That is a fault directly traceable to a design (and probably the management) from that company.

Learn about a device called a voltage supervisor. Some shave costs by eliminating this sophisticated circuit or by duplicating it with a circuit that was even piss poor in the 1970s. Its not power loss that causes these problems. Problem is bad product designs.
Last edited by westom on Jul 15th, 2017 5:12 pm, edited 1 time in total.
Member
Jul 6, 2009
384 posts
110 upvotes
willilumplump wrote: Unit reports the battery is fine and everything is hunky-dory. It's only a year old, so I wouldn't expect a problem.
Plenty of technical reasons can explain that. For example, a UPS must specify (with a number) its switchover time. That computer must define how long it can still operate with no incoming AC power. Yes, all electronics will operated uninterrupted without power for a given time. A minimum is about 17 milliseconds. Better electronics routinely operate without power even for 100 milliseconds.

A power outage benchmark is a VCR or microwave oven clock. These tend to detect a power loss (end up blinking) much more easily than anything else. If that clock did not detect a blackout (or longer brownout), then power loss probably did not happen.

View the UPS. These things are often designed as cheaply as possible. Battery life expectancy is only three years. Noise and other anomalies can trigger a UPS into thinking a blackout has occurred. Some even get confused when power cycles on and off, or when voltage keeps rising and falling between a fixed threshold.

Possible that a UPS kept assuming power was there when in fact it was not. Or worse, got confused and switched off both AC and battery power due to that confusing power cycling.

That same 'confused' power cycling is why better computer systems have a protective lockout. Rather than trying to decide what is and is not good power, a computer simply locks out power. That lockout gets reset by disconnecting its power cord for less than 10 seconds.

(Some use speculation to assume this is letting a computer discharge. It does not. Power cord is disconnected to reset a hardware safety function.)

This strange UPS behavior is why a manufacturer, model number, and relevant specification numbers are necessary to obtain a better informed reply.

Same applies to a battery. How does a UPS know that battery is fine? A defective battery can still measure good. The fact that the battery did provide power for sufficient time implies a good battery. However, was that battery fully charged? A UPS is made as cheaply as possible. A power source that does recharge is made as tiny as possible. So a UPS battery often needs at least 12 hours and sometimes 24 hours to be recharged. Was that battery sufficiently charged to support a small load, but not a full load? Again, an informed answer means perspective - that means numbers.
Deal Expert
User avatar
Mar 14, 2009
16761 posts
8093 upvotes
I really like this @westom guy. Extremely knowledgeable and tells it like it is. I find it comical watching you guys downvote him for no good reason.
Deal Expert
User avatar
Mar 6, 2003
19712 posts
13576 upvotes
Ottawa
SickBeast wrote: I really like this @westom guy. Extremely knowledgeable and tells it like it is. I find it comical watching you guys downvote him for no good reason.
He offers no advice of practical value. Power outage of your computer? Just reboot. Everything will be fine, you just lost the little bit of unsaved data. No need for a UPS, modern hardware and hard drives and OSs handle preserving your data

Ya right :facepalm:
UPS provides temporary and 'dirty' power so that unsaved data can be saved. And to avert an inconvenient reboot.
Nothing more.
Please update your profile to include your city https://forums.redflagdeals.com/ucp.php?i=ucp_profile&mode=profile_info
Deal Expert
User avatar
Mar 14, 2009
16761 posts
8093 upvotes
warpdrive wrote: He offers no advice of practical value. Power outage of your computer? Just reboot. Everything will be fine, you just lost the little bit of unsaved data. No need for a UPS, modern hardware and hard drives and OSs handle preserving your data

Ya right :facepalm:
I've been using computers for over 20 years and I have never once lost data due to a power failure. Not one single time.
Deal Expert
User avatar
Mar 6, 2003
19712 posts
13576 upvotes
Ottawa
SickBeast wrote: I've been using computers for over 20 years and I have never once lost data due to a power failure. Not one single time.
good for you. Sample size: 1
Please update your profile to include your city https://forums.redflagdeals.com/ucp.php?i=ucp_profile&mode=profile_info
Deal Expert
User avatar
Mar 14, 2009
16761 posts
8093 upvotes
warpdrive wrote: good for you. Sample size: 1
Do you get frequent brownouts where you live? I don't. Maybe once or twice a year. If you get a lot of power failures it could be worth it, or else if you are doing a lot of important work on your computer where your data is super important. I personally backup everything critical automatically to the cloud and take weekly incremental backups of my main drive. I don't do critical work. So I don't think twice about having a UPS. You can't paint everyone with the same brush. We all have different usage patterns and needs. @westom has a perfectly valid opinion IMO.
Deal Expert
User avatar
Mar 6, 2003
19712 posts
13576 upvotes
Ottawa
As a guy who have been working on computers for a living (and mainly high power desktops) for longer than you have. I would say having a UPS is good insurance for both unsaved work and the integrity of the disk drive contents. 99% of the time, you won't need it. There's a difference between saying that and saying what Westom is saying.

He's saying things like, oh yeah, that's old thinking. No longer a problem. It's a myth perpetuated by naïve or stupid people. It's a false premise. it's junk science because mine is real science.

He already made some false statements with regard to data loss. Look at SSD report which is based on real testing. Look at people reporting nonbooting computers. All his theories account for nothing. It's the junk science. He's probably the first guy to proclaim that the Autopilot system in the self driving car cannot fail, or if it fails, there will be little consequence. Modern sensors, algorithms, and object tracking is robust now he says, that was a problem in the past. And then somebody dies in a high speed head on crash with a stopped object. You can be knowledgeable but still be way off base in a real world scenario.

No I don't live in a brownout prone area, but I have had severe down time due to an unforeseen power interruption.
Please update your profile to include your city https://forums.redflagdeals.com/ucp.php?i=ucp_profile&mode=profile_info
Deal Expert
Aug 2, 2004
38392 posts
12014 upvotes
East Gwillimbury
SickBeast wrote: I've been using computers for over 20 years and I have never once lost data due to a power failure. Not one single time.
I hope you never have a power outage while working on a spreadsheet or a Word document.
Member
Jul 6, 2009
384 posts
110 upvotes
warpdrive wrote: good for you. Sample size: 1
Good for me. Thousands. Experience not from junk science; I even learn from datasheets.

Good for you. "I think a brownout does damage. I do some things that I don't know what they do. Then symptoms are cured. So I am an expert."

Engineers witness this ignorance often. Not one engineer said it was safe to launch Challenger. It was launched by the naive who routinely ignore numbers and who claim they know better using subjective reasoning. Boisjoly whispered to Ebeling, "we just dodged a bullet," Seconds later, Challenger exploded - as numbers said it could. According to Ed Corrigan (Christa McAuliffe's father), people were killed because of "egos, marginal decisions, ignorance and irresponsibility. "

Same exists here. Egos, marginal decisions, technical ignorance, irresponsibility, and knowledge without any numbers have repeatedly denied reality. So many computer repair techs have no idea how a computer really works. Much computer repair is done by self proclaimed 'experts' who only understand shotgunning. Who even foolishly assume brownouts are hardware destructive. Technical ignorance is apparent when accusations and denials never posts any datasheet or specification numbers. So many just know due to egos, marginal decisions, ignorance, irresponsibility - and not one number.

A consumer magazine created defects in tens of computer. Then took them to computer repair shops. A majority were returned without fixing the defect. Most computers got a perfectly good PSU replaced. Technical ignorance because so many felt a PSU must be replaced. First indication of the most technically naive? Feelings and experience prove it must be true. Not one number that said a PSU was defective.

"I drive a car. So I am an expert on how car motors, brakes, and engines work." Same experience also proves a UPS or adjacent protector protects hardware. Bogus (subjective) reasoning somehow knows better. Neither do effective protection. No facts with numbers are provided. Instead, "I know a UPS protects hardware because feelings (classic junk science) are proof."

lostintransit asked for effective solutions. We never got there due to so many nasty denials, bogus reasoning, outright lies, too much ego, and no numbers. No UPS protects hardware. No UPS number even claims to protect hardware. UPS is temporary and 'dirty' power to save unsaved data. 'Dirty'? Yes, because best protection, already inside electronics, makes 'dirty' power irrelevant. Electronics are robust - already contain superior protection.

Concern is for anomalies that can overwhelm that best and existing protection. That something is completely unknown to many computer repair techs who post denials. That solution is always found in facilities that cannot have damage - even 100 years ago. That best solution costs about $1 per protected appliance. That solution is just what the OP and lostintransit are asking for. But we never got there due to so many poorly trained computer repair techs that do not know what that something is.

Plenty of numbers and installation guidelines exist for best protection. But the nasty and naive want to argue. As if shotgunning makes one an expert.

Not one engineer said it was safe to launch. So self proclaimed experts used ego, marginal decisions, technical ignorance, and knowledge without any numbers. Same types obtusely insist that brownout and blackouts harm electronics. Tiniest joules in a UPS does not protect hardware.

Over 100 years of well proven protection, that costs about $1 per protected appliance, is recommended to protect all household appliances. An engineer who even did space flight hardware could have provided the OP and lostintransit with informed answers. But technical ignorance, posted by self proclaimed experts, make honest discussion impossible.
Deal Expert
User avatar
Mar 6, 2003
19712 posts
13576 upvotes
Ottawa
SickBeast wrote: I really like this westom guy. Extremely knowledgeable and tells it like it is. I find it comical watching you guys downvote him for no good reason.
Well, he's going on my ignore list. Too much grandstanding and too little (actually: none) proof to back up some of his claims. It's ironic that he claims things are false because there are no numbers yet he has yet to actually back up his own claims with data or credible sources.
Please update your profile to include your city https://forums.redflagdeals.com/ucp.php?i=ucp_profile&mode=profile_info
Member
Jul 6, 2009
384 posts
110 upvotes
warpdrive wrote: It's ironic that he claims things are false because there are no numbers yet he has yet to actually back up his own claims with data or credible sources.
Does the expression "specification numbers" mean no numbers?

Rather than be nasty, why not post that number and ask for clarification. One does that when one wants to learn. So far, you post accusations - and no numbers.

No numbers? 20,000 amps, 100 years, 17 milliseconds, less than 10 feet, 200 volt square waves with 270 volt spikes, hundreds or thousand joules, microsecond transients, low impedance, how protection was done 250 years ago, $1 per protected appliance, 15 million protectors that are a serious fire threat, a 2 cm part that somehow blocks what 3 miles of sky cannot, well over 90% of all lightning strikes cause no apparent damage, bulbs can dim to 50% intensity, 120 ohms impedance creating 12,000 volt transients, numbers and charts from ignored international design standards, 1800 volt internal protection, what fuse specs for milliseconds and 250 volts report, and so many companies known by any guy for integrity. Unfortunately eyes glazed over with numbers. So you never saw any of them - to protect well entrenched fables.
Deal Fanatic
Dec 12, 2009
6145 posts
3614 upvotes
Toronto
westom wrote: Concern is for anomalies that can overwhelm that best and existing protection. That something is completely unknown to many computer repair techs who post denials. That solution is always found in facilities that cannot have damage - even 100 years ago. That best solution costs about $1 per protected appliance. That solution is just what the OP and lostintransit are asking for. But we never got there due to so many poorly trained computer repair techs that do not know what that something is.
I am having a hard time reading through these circular monologues trying to figure out what is being said.

@westom can you say what this $1 solution is in 5 words or less? How about 1 word?
Member
Jul 6, 2009
384 posts
110 upvotes
ROYinTO wrote: @westom can you say what this $1 solution is in 5 words or less? How about 1 word?
If anything does not also say why, then the recommendation is ineffective - probably a lie.

Demonstrated is why plug-in protectors and UPS manufacturers do not claim to protect hardware from potentially destructive surges. Those are classic one word or soundbyte recommendations. If one does not know why a "$1 per" solution works, then it cannot be implemented. Effective solution involves an installation and an interconnection of multiple devices. Effective and many times less expensive solution is a 'system'; not a magic box.

An effective solution is described IF really desired. If one wants a soundbyte or one word reply, then one wants a magic box; not an effective solution. Magic is why so many spend massively on a near zero joule UPS or power strip ... that does not, does not claim, and cannot provide effective protection.

Effective solution is layman simple. But many completely ignored concepts apply. A solution well understood even 100 years ago takes at least four paragraphs just to introduce. I learned long ago that anyone who cannot read beyond a few paragraphs only want a magic box; do not really want a solution.

Do you want that well proven solution?
Deal Fanatic
Dec 12, 2009
6145 posts
3614 upvotes
Toronto
ROYinTO wrote: @westom can you say what this $1 solution is in 5 words or less? How about 1 word?
westom wrote: Do you want that well proven solution?
I'd prefer you answer the question.
Deal Expert
Aug 2, 2004
38392 posts
12014 upvotes
East Gwillimbury
ROYinTO wrote: I am having a hard time reading through these circular monologues trying to figure out what is being said.

@westom can you say what this $1 solution is in 5 words or less? How about 1 word?
Based on his recent reply. It appears the answer is no

Top

Thread Information

There is currently 1 user viewing this thread. (0 members and 1 guest)