Root Access - 2015-07-21 (page 1 of 3)

« first day (1810 days earlier) ← previous day next day → last day (3224 days later) »

12:20 AM

@DragonLord I have no love for Verizon at all. They told me in 2007 that it would be "weeks" (quoting a rep, there) until I'd have FiOS. It's been several hundred weeks and I'm still waiting. Apparently they forgot about our community because it wouldn't make them enough $$$ to bring us FiOS, so it's only available like 1/8th mile away.

ISPs will not get my respect until they stop trying to milk the world of every penny they have, and get out of the way of consumers and businesses trying to build a world on top of the Internet. Right now they are an impedance, just like the power companies were in the early days of electricity. It took something ridiculous like 60 years to make power companies behave.

Then it took several decades again to make phone companies behave. Every time a new industry is created, the first guys there always want to milk it while delivering as little as possible - basically they're always run by assholes.

The default mode of treating other humans in a capitalist society is equivalent in every way to totalitarianism.

@User95050: unless gravity suddenly goes out of whack, floor tiles cannot possibly fall on the tub. — Martha 3 hours ago

@allquixotic ISPs? Ha. Try the governments (too).

(Though, in all fairness the US fed gov seems to have at least tried, and apparently some local govs are setting up fibre networks.)

@Bob Ha

@Bob Well, yeah. But your governments are being rudely "captured" by corporations, so that they're basically one and the same.

12:38 AM

@allquixotic And yours isn't? :P

Also, when I said local govs I meant US local govs.

AU local govs are doing jack all.

2 hours later…

2:12 AM

Hey, I just found @MichaelKjörling on oldnewthing :P

2 hours later…

Journeyman Geek

3:56 AM

lol

so that's why I keep changing my handle every decade or so ;p

@JourneymanGeek I changed from my first handle cause 'the_bomber_man_2001' stopped being appropriate...

Also, cause I was no longer 13 :P

@MichaelFrank O_O

now that's a username to cause arguments

@Bob Yeeeeeep. To be fair, I had just started intermediate school, and we must have been playing a lot of Bomberman at the time.

Journeyman Geek

lol

good game.

Lets see, blizzard of brick (In my lego fanatic stage. I miss lego, but I don't have the space for proper mucking about), geek and other RPG charecters I played in the past.....

Journeyman Geek

4:28 AM

I need to remember to cancel one of my VPSes when I get the time. So darned busy T T

4:42 AM

@JourneymanGeek I still need to move my server and cancel the old one!

That's 70 USD per month O_O

should prooooobably do that ASAP

Journeyman Geek

lol. this is ~23USD a year. and I can cancel the database too. Prolly keeping the super cheap vps tho

@Bob: heh, most of my necessary moves are off stuff I don't pay for, or run at home

I need to move znc, and the stuff I run on allquixotic's server. Tho, it I could make sense of multimaster postgres clustering, tempted to try running ttrss on top of that, and having it as a HA setup ;p

o.O

and here I am procrastinating on kvm

Journeyman Geek

lol

I'm procrastinating on most of these things. And ipv6 for the new server

(tho the ipv6 setup I want will prolly need an SU question. Pity Its not SF's scope, even if its a dedi. )

5:04 AM

@JourneymanGeek depends how you ask it

Journeyman Geek

5:21 AM

True. But its less enterprisy than wierd and badly documented :p

Journeyman Geek

5:32 AM

But been busy with the new monitor/setup the last few days so.... ;p

5:42 AM

@JourneymanGeek still waiting for my RAM :(

Journeyman Geek

lol

currently in the saving/planning phase ;p

>
Your package is delayed in the customs clearance process.

Journeyman Geek

/me is still not decided on his new GPU

And I'll buy it locally so, semi instant gratification ;p

Expected to arrive next Wednesday.

Journeyman Geek

Man. I've NEVER had that here

5:44 AM

1 AUD = 0.73 USD ... yea it went down a lot

Journeyman Geek

tho, if its >400 dollars it sometimes gets detained by customs, and you need to pay at their office and pick it up, or pay singpost to do it for you.

Mine is only something like 150 USD.

The threshold for GST is 1000 AUD, I think

unless they changed it

1 AUD = 1.01 SGD

So, yea, parity there :P

Journeyman Geek

Oh, here you don't have GST for overseas purchases under 400sgd

but meh. I rarely buy stuff online that's that pricy. Most of it is random bits off dx or such

Oh and clothes, can't always find pants that fit here

o.o

they don't sell dog pants in Singapore?

Journeyman Geek

6:00 AM

._.

No? ;p

None that fit anyway ;p

5 hours later…

10:48 AM

sigh

CloudFlare's Sydney network is currently all but down

11:29 AM

@Nick Less clear but more attracting title and still same topic :

area51.stackexchange.com/proposals/89327/…

12:11 PM

github.com/robertdavidgraham/masscan

2 hours later…

Oliver Salzburg

2:09 PM

One of those moments again...

I've been wondering why my Nagios graphs no longer contain GPRINT directives

And suddenly I find this:

1

Q: Why are my GPRINT directives not being drawn into my graphs?

I'm using NagiosGrapher 1.7.1 with our Nagios installation. We're using the packages from the official Debian repositories. The graphs I've been building work fine so far, except for any GPRINT directives I use. They're simply not processed at all. An example of a graph definition would be: de...

debian-wheezy rrdtool nagios

Journeyman Geek

;p

3:08 PM

So I'm pretty annoyed that right now Windows 8.1 is resisting all attempts to fix Windows Update.

o.O

What's wrong with it?

And didn't MS release some repair tool a while back?

Oh here it is windows.microsoft.com/en-au/windows/…

3:22 PM

sfc /scannow can't fix corrupted files. dism /Online /Cleanup-Image /RestoreHealth fails because Windows Update is broken. dism /Online /Cleanup-Image /RestoreHealth /Source:wim:C:\win81\sources\install.wim:1 /LimitAccess tries to clean up from an unpacked iso with the .esd converted to .wim, but it fails because it claims not to be able to find the source media. the StartComponentCleanup flag returns Operation completed successfully, but it doesn't actually fix anything.

@Bob the Windows Update repair tool says it fixed stuff, but even after rebooting, Windows Update still hangs indefinitely upon trying to install even the most trivial update.

I tried repairing both BITS and Windows Update itself.

o.O

@allquixotic hate to ask this, but have you checked the logs?

@Bob not sure what logs to check - the sfc scannow logs? the DISM logs? not sure where Windows Update logs would be.

@allquixotic support.microsoft.com/en-au/kb/902093

might help to know which step it's failing on

rather than guessing it's BITS or other

I should probably also do a RAID array integrity check while I'm at it - our mains have been relatively unstable of late (as in, not constant brownouts, but we've lost power enough to turn off the PC about 6 times in a month)

@allquixotic you need a UPS

3:26 PM

@Bob don't remind me. I have my sights set on a 1500 VA Online Pure Sine Wave PSU, but I can't afford it right now

that MS KB article is actually amazingly thorough

this thing is incredible:

@allquixotic which one?

amazon.com/CyberPower-OL1500RTXL2U-Online-1500VA-100-125V/dp/…

O_O

3:26 PM

it's not line-interactive at all

@allquixotic that doesn't make much difference, really

pretty sure I've had this discussion in here before, too

line-interactive switchover is fast enough that the vast majority of PSUs will never notice

@Bob my Corsair PSU is pretty damn sensitive to tiny perturbances that barely make the lights flicker, and things like alarm clocks, microwaves and fans are like "what? did something happen?"

@allquixotic If you notice it, it's not tiny.

line-interactive switchover is something like 5ms

Human-noticeable blips are generally at least an order of magnitude larger

still, I'd rather not skimp on UPS quality, and an online one is probably the best in breed

Jun 15 at 15:19, by Bob

@DragonLord again, negligible benefit. it's better on paper, but in practical situations you won't notice a difference apart from extreme edge cases

@allquixotic even then, not really!

Jun 15 at 15:30, by Bob

@DragonLord You might find this APC white paper interesting.

3:30 PM

that one is also upgradeable with extended battery packs if I ever decide I want them, and you can replace the primary battery pack - I'm sure there are some line interactives that let you do that too, but meh

if you click through and read from there, that entire discussion was basically discussing exactly this

anyway, it's not like I'm trying to spend $5000 for a 1500 VA power supply - at most it's $200-$250 more than an equivalent line-interactive PSU, and some line interactives are just as expensive as this one

I can't get anything higher than 1500 VA, either, because then you switch over to the NEMA plugs that demand 20 Amps mains input, and I don't have any of those on the top floor of my house and I'm not willing to install one just for a PSU

> SMPSs are the power supplies found
in virtually all mission-critical equipment. They
must have a “ride-through” time of 10 milli-
seconds or more to comply with international
standards (see Figure 3).

(we have 20A outlets in the utility room in the basement)

@allquixotic We just use 10A @ 240V :)

> Advantages of line-interactive topology
:
•
Lower electricity consumption (less costly to
operate) – More efficient because less power
conversion is performed when acceptable AC input is present.
•
Theoretically higher reliab
ility – Lower component count
and lower operating temperatures. (See later section
Reliability Considerations
.)
•
Less heat load on the facility
– Less heat is produced by
the UPS.

3:33 PM

@Bob heh. standard is 15A @ 120V. most/all houses built since the 1970s in the US also have at least one 20A receptacle somewhere, but they're usually for things like refrigerators, (electric) stoves, and washing machines.

> Things to watch out for:
The line-interactive UPS may not be the appropriate choice for
installations where
•
AC power is unstable or highl
y distorted, because battery
power will be used too oft
en to keep the UPS output
within specifications.
•
Power Factor Correction (PFC) is required and the load
equipment does not perform this function.

You don't really need to worry about PFC in a home environment, and your PSU does it anyway.

Your power (8x a month) is not classified as unstable by their standards.

They're talking multiple times a day.

@allquixotic We go 15, 20, 25A for the really beefy things, but you won't find them in most homes.

Maaaaaybe some (older) resistive electric stoves.

@Bob I'm also not sure if my power has a tendency to drop below the rated voltage - I don't think it does; generally it's either on, or briefly tripping off (for between a few tens of milliseconds to several seconds for more significant problems). I'd imagine the US power standards are probably pretty good at preventing extreme undervolt conditions?

> MYTH
: On-line UPSs offer better protection
against Common-Mode (CM) noise.
REALITY:
Even though it is possible to design
both on-line and line-interactive topologies with
galvanic isolation, they both typically use passive
components to
reduce
CM voltage. Neither
on-
line
nor line-interactive topology offers a
fundamental advantage in this respect. SMPSs
already have galvanic isolation, so external
isolation is unnecessary. See APC White Papers
#9 and #21 for more information.

@allquixotic A slight drop is fine. It's only the really big drops that cause issues. And they're generally very rare in developed countries.

(We did have two ~120V brownouts a year apart, lasting several hours... that was a couple years ago now)

@Bob if it gets that low, wouldn't the transformers or whatever at the mains source just turn the power off?

(for safety)

@allquixotic Depends how they're designed and set up.

3:36 PM

My Roccat Tyon seemed to have experienced a glitch (likely in firmware) this morning.

Often they'll keep them on. Why? I don't know. Maybe so people still have light?

The issue seems to have sorted itself out after a firmware update, driver reinstallation, and reboot.

They did kill two of our laser printers, though.

They paid the replacement costs.

@allquixotic I'm talking an actual configured low-voltage mode.

I don't think it was a coincidence we went from 240 to 120.

It affected one of the auxiliary buttons (the upper right one), causing it to not respond unless I hit Easy-Shift[+], where it then becomes stuck on.

A sag on the line is somewhat different.

@allquixotic Anyway, gist of what the UPS manufacturers themselves (APC, TrippLite, the big ones) say is: don't bother with online at home unless your power is ridiculously bad.

3:39 PM

The problem cleared up by the time I rebooted into WinRE in an attempt to diagnose the issue.

Costs more both short and long term, too (efficiency losses, need to replace parts more often)

@Bob What needs to be replaced more often? The power electronics, or the batteries?

@allquixotic Both.

> Double-conversion on-line contains
more components that run conti
nuously at higher temperatures
and, with all other things being equal, have less servic
e life than similar parts
found in line-interactive.

Primarily the power electronics, but the heat will adversely affect batteries too.

> For line-interactive, the small number
of components and cool operati
on of the main power stage
both tend to
increase
operating life and reliability. For double-
conversion on-line, c
onstant operation and
higher operating temper
atures both tend to
decrease
operating life and reliability.

@Bob 240V AC power has its own idiosyncrasies. Here in the US, the 120V grid doesn't malfunction in the same manner.

There is the caveat:

> In practice, however, reliability is generally determi
ned by how well the manufacturer designs and builds the
UPS and by the quality of the com
ponents used, regardless of topology.
Because quality is vendor-dependent,
there can be high quality double-conver
sion on-line designs and poor quality li
ne-interactive designs, and vice
versa.

@DragonLord That has nothing to do with 240 vs 120 V

That would just be what the local power company did.

There is nothing inherent to the line voltage that would define how a brownout looks.

3:42 PM

@Bob This complexity is compounded by the fact that money doesn't always equal better. A shoddily built product can be priced high, and a very efficient manufacturer can produce great quality stuff at reasonable prices.

(Except that a 240 V brownout is more likely to let some SMPSes continue to operate - our networking equipment stayed up through the last two.)

for the first example, look at Beats

for the second, Meelec ;p

@allquixotic As a rule of thumb, you basically gain no advantage from online and you lose out on efficiency if nothing else.

Unless your power is extremely dirty.

@Bob It sounds like that's more true In the US (and other such developed countries) than, say, Pakistan.

A while back I was looking into a double-conversion UPS:

Jun 15 at 15:14, by DragonLord

This is the UPS I want for my future PC build: http://www.newegg.com/Product/Product.aspx?Item=N82E16842111478

Jun 15 at 15:14, by DragonLord

Online double-conversion 1500VA / 1350W.

3:44 PM

@DragonLord Yea, the stuff I linked to above was basically from that conversation :P

One of the reasons for this choice was the much higher wattage rating.

The line-interactive model (Tripp Lite SmartPro) is good for only 900W despite the 1500VA rating.

Bollocks. The only relation between watts and VA is the power factor.

On a system with active PFC, there might be a problem under very high load.

Especially with active PFC (and pure-sinewave line-interactive), I suspect they'd be nearly equal.

tripplite.com/…

3:47 PM

The entire point of PFC is to bring the power factor as close to 1 as possible, such that W ~= VA

^ this is the line-interactive equivalent.

I'd be limited to the 900W number.

@DragonLord Apparently because it's not available in my country I can't even look at it.

You're gonna have to screenshot it.

There were some discussions about a gaming PC potentially needing more than that, but I guess those were overblown.

@DragonLord No you wouldn't.

UPSes are rated in VA. Period.

The W number is the manufacturer's guess, and the relationship with VA will depend on the PF.

Where the PF = Watts / VA

> SMPSs with passive PFC can achieve power factor of about 0.7–0.75, SMPSs with active PFC, up to 0.99 power factor, while a SMPSs without any power factor correction have a power factor of only about 0.55–0.65.

That would be where the 'normal' PF guesses for UPSes come from.

900 W would be assuming a PF of about 0.6.

@Bob So is the power factor based on whether or not the device using the power downstream has active PFC?

3:51 PM

@allquixotic Yes.

@Bob Example. Take a line-interactive 1500VA pure sine wave PSU. Separately, and after a full charge, plug in a PC with a great active PFC power supply. Then repeat the test anew with a PC whose PSU has no power factor conversion. Will the second one drain the batteries faster?

@DragonLord I notice that that UPS does not specify that it is pure-sinewave.

user image

Active PFC has been reported to have issues with the simulated sine wave typical of traditional line-interactive units.

@Bob I was wondering if "PWM sine wave" is going to be a problem.

It's not quite the usual simulated sine wave, give me a moment to pull it up

3:54 PM

@DragonLord @allquixotic I'm not too comfortable with recommending purchases based purely off my understanding of how this works. But I am reasonably certain that this is correct.

@DragonLord Well, "pure" isn't really pure.

They normally mean better than usual.

This might be the better solution: amazon.com/APC-SMC1500-Smart-UPS-Interface-Uninterrupted/dp/…

Pure sine wave output.

So now I just need to find a power cord for my computer that is long enough to run across the room to the UPS. My room gets hot enough as it is; I don't need a ~~heater~~ UPS in there at the same time.

@allquixotic My UPS is barely warm.

Actually, it's not warm at all.

Probably because it's basically asleep.

Meanwhile, an online UPS is probably losing 5-10% of the power flowing through it as heat.

apc.com/products/resource/include/…

Mine has a minor warm-ish (i.e. not freezing like the current room temp) patch at the back.

At the end of the day: online would certainly work, and is less likely to have compatibility issues with equipment. However, you pay more both at the start and during their lifetime, for questionable benefit.

On the other hand, a modern line-interactive unit (rated/tested/guaranteed to work with active PFC!) is unlikely to have issues anyway, and is cheaper.

If you were to pay the same amount, you could probably get a better (higher-end) line-interactive.

As for rated VA vs W, with active PFC the circuitry should be able to handle W near VA. However, there are two possible issues: a lower W rating might imply a smaller battery, and they might have already accounted for the discrepancy with some PFC on the output side of the UPS (in which case I question the implicit assumption of a PF of 0.6...).

Unfortunately, there is little to no discussion on this anywhere that I can find.

So basically I'm making this up as I go along based off the definitions of PF, PFC, W, VA and what I can find individually on those :P

(i.e. I'm not really qualified in that area.)

...come to think of it, I actually know a guy who is qualified there. Maybe I should just ask him O_O

> APC has addressed the issue of running Active PFC PSUs on their line of consumer, stepped sine-wave output UPSs. Basically, their advice in such cases is to size the UPS according to the maximum rated output of the Active PFC power supply (due to potential power rushes), not according to the peak power consumption of the PC.

@DragonLord @allquixotic ^

Another potential issue. Peak (inrush) current draw when powered on, or bouncing back from the 5ms switch.

That's where online might perform better.

But a single PSU isn't going to draw enough for that to be an issue...

More interesting stuff: emersonnetworkpower.com/en-ASIA/Products/ACPower/NetworkUPS/…

> Liebert GXT3 rated output power factor of 0.9 better matches up with switch-mode power supplies used in today's IT equipment. A higher output power factor allows for a more efficient utilization of the UPS.

So here we have a non-1 PF output from the UPS o.O

Seems like they did that intentionally under the assumption that the average computer has a PF < 1

i.e. they do some form of (probably-passive) PFC internally (?)

The "more efficient utilization" seems to confirm my assumption about active PFC on the PSU (bringing it near 1) helping, assuming most UPSes are expecting (rated according to) an optimal PF of 1.

Heh, looks like @Sathya made the same assumption: superuser.com/questions/45138/…

4:20 PM

ahh, I just learned of a potential use case for online PSUs

Though, reading a bit more, it looks like the watt rating on some UPSes might also determine the heat dissipation capability... that could be a problem.

Safest bet is just to not exceed either rating.

@allquixotic Hm?

diesel or gasoline generators sometimes start-up and provide very dirty power during the early milliseconds to seconds of supply, then level out somewhat - but during that beginning stage, it might be ugly enough to disturb a line-interactive system

Ah.

apc.com/resource/include/techspec_index.cfm?base_sku=SMT1500

@allquixotic Do you mean disturb as in force them to switch to battery (which is what the white paper above was talking about, too much switching) or cause them to cut out?

I vaguely recall hearing of the latter.

4:22 PM

@Bob the latter - not so much that the UPS couldn't handle it, but the power delivered to your sensitive PC would cut out and your PC would turn off :P

the UPS would be fine, in the end, and so would your PC, but you'd lose whatever work you were working on at the time on your computer

SMT1500 is more expensive than SMC1500 but has higher watt rating (1000 W) and wider input voltage tolerance

@allquixotic :A good UPS should detect and switch on dirty input power.

True, by their nature an online UPS should never let that through.

Whether a line-interactive does or not would depend on how well it's designed.

Runtime is also longer

I guess I'll use this model for my next PC. 1000W is plenty.

we are planning to possibly install a natural gas generator supplied by the residential gas pipeline, since we live in an area that is on the actuary tables for worsening hurricane, thunderstorm and tornado risk over the next 25 years

@DragonLord Interesting tidbit: that's actually the UPS (or very similar to) that we're running a T620 server on :P

@allquixotic That sounds... pointless.

If it's supplied by utilities that can be interrupted in the event of a disaster.

A generator with an independent fuel source (usually diesel) sounds more useful in such an event.

4:26 PM

@Bob howso? AFAIK the gas keeps flowing when power is lost, so it would take something like an earthquake (which we're not at especially high risk of) to damage the gas pipelines

What's the chance that you'll be without power but with working gas?

@Bob Depends on the nature and extent of the disaster. If power lines are down because of lightning or a localized tornado, we'll still have gas.

If it's a category 5 hurricane affecting the entire area, yeah...

@allquixotic Lightning tends to be a pretty quick fix.

Might be able to get one that relies on the gas pipeline if available, and if not, we just store a barrel full of boom (KSP reference) in the backyard

Tornado... not so much, but they can often reroute around the affected area.

Assuming you aren't actually directly hit.

But gas itself... what's your source?

Is it piped in?

Shipped? Trains? Trucks?

Unless it's piped from all the way outside the potential disaster area, gas could be affected too.

And most large pipes are also above-ground...

I dunno.

4:31 PM

@Bob Actually, that's complicated. Residential natural gas and electricity are both deregulated in my state, where you have the supplier and the infrastructure owner. Constellation Energy is the infrastructure owner, they own the pipes that run from my house to wherever the gas comes from. Then we selected a supplier who is responsible for getting the actual gas and delivering it through the pipes.

I imagine there are multiple sources, including trains, boats, and probably a pipeline to some other area.

@allquixotic In that case, it would be to a reservoir somewhere.

I believe (don't quote me on this) that the majority of our gas comes in huge bulk quantities through the Port of Baltimore on barges.

Then, question is, how likely is a disaster to knock out your power while leaving the gas unaffected, knowing that they both have similar weak points?

They buy it from wherever it's gathered (e.g. in the same place we get crude oil, usually) and order it in large batches to be stockpiled then delivered.

(you need to ship in resources, you need to generate/store it somewhere...)

4:34 PM

@Bob Again, depends on the severity and duration of the disaster. Most of that industrial equipment is going to be very resistant to actual physical damage from any weather related phenomenon (excluding major geological events), but if there's a long-term (weeks) shutdown of the Port of Baltimore, we may lose gas even if we still have power from the nuclear station.

if you're planning for the eventuality where you're physically untouched but power goes down (i.e. disaster is localised to other area or you survived a massive disaster)... probably better to make sure you aren't completely relying on utility gas

You aren't going to see natural gas tanks exploding because of a hurricane. They're very sturdy.

@allquixotic Not exploding, but a large storm could potentially flood out the pumping facilities?

Maybe.

Just throwing random (pessimistic) possibilities out there.

But as long as you're preparing for a disaster, might as well make the prep complete as possible.

Yeah. Like I said, the most robust solution would be to have a tank of gas on-premises that (if rationed) could last us a given period of time, plus also connect to the pipeline.

That would also allow us to automatically "refill" the on-site gas reservoir in case it gets used up. They'd probably install an automatic valve to do just that.

Or a manual one *shrug*

4:53 PM

Is "Root Access" the name just for the computer reference or is this chat room only about root access?

@SarahofGaia It's just a name. This is the "official" (ish) chatroom for the Super User Stack Exchange Q&A site.

Ah, I see. Thank you. :)

Don't think of it as being about any particular topic. The actual room may as well be called "Chatroom for people who are interested in the Super User site".

Oh lol ok

Could you point me in the right direction for something?

!!tell 22919648 meta

4:59 PM

XD

Damn, Cavil is still down :P

What is 22919648

@SarahofGaia Sure. What do you need?

What? o.o I'm confused.

I thought you were referring me to another place...

@SarahofGaia Nah, that was just a chatbot command.

4:59 PM

ohhhhh

okies

so anyway

« first day (1810 days earlier) ← previous day next day → last day (3224 days later) »

Transcript for

Jul '1521

Root Access

For all you Super Users out there. You have backups, right?

computers technology

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile