NASA resets computer, repairs data corruption on Voyager 2 interstellar probe | Hacker News Reader

NASA resets computer, repairs data corruption on Voyager 2 interstellar probe

jpl.nasa.gov

135 pointsterra_t48 comments

Whenever I read stories like this, I always look in awe at the white-beards in my building who were around back in the days when computers were machines instead of abstractions.

The Story of Mel -- A Real Programmer comes to mind:

http://www.pbm.com/~lindahl/mel.html

Go program a microcontroller and you'll grow a grey beard all your own.

http://blog.jgc.org/2009/08/just-give-me-simple-cpu-and-few-...

The Voyager computers are of similar scale to microcontroller.

This is really amazing given the age of the Voyager craft and the massive distances they have to beam instructions across. I'd imagine that all of the engineering staff would be on pins and needles for hours while they waited for what has to be the galaxy's largest (in terms of distance crossed) SYN-ACK :p

As someone who works with autonomous robots, restarting a system 30 feet away in demo conditions can cause pins and needles. I can't fathom how the NASA engineers felt. They probably didn't get any good work done waiting for the ACK

It's incredible that NASA, 30 years ago, built such a durable and reliable system. The hardware has withstood one of the harshest environments in the universe for decades, and the software hasn't been irreversibly scrambled by passing cosmic rays. My cable modem doesn't even last an afternoon without rebooting itself.

The Egyptians built the pyramids 5000 years ago. What I find incredible is how we can't seem to build things that last anymore.

Market forces - people are quite happy paying for cheap stuff that doesn't last very long.

Yeah, but they built them for religious reasons, not because they actually wanted to do something useful with them (for some definition of useful).

Sure, we could build pyramids – better pyramids even than the original ones – but they wouldn’t be terribly useful.

(Oh, and sometimes we do build things that last. The six Apollo descent stages are still on the moon and probably will be for Millennia. That, to me, is a lot more impressive than any pyramid.)

> Sure, we could build pyramids - better pyramids even than the original ones - but they wouldn’t be terribly useful.

There are innumerable Discovery Channel-esque documentaries explaining that we can barely build a small scale reproduction of the Pyramids, at least using the knowledge they had available then.

I think there are still some attempts to build things to last:

https://secure.wikimedia.org/wikipedia/en/wiki/Svalbard_Glob...

We build things intentionally so that they don't last long anymore. There's an entire realm of engineering dedicated to predicting failures of mechanical devices at a certain range. The ipod comes to mind of course, but there are many more all designed to maximize returns by having customers buy new ones.

http://www.longnow.org/

http://en.wikipedia.org/wiki/Deep_geological_repository

http://en.wikipedia.org/wiki/SubTropolis

Las Vegas will last forever. Arid climate, Lots of concrete.

On the bright side, I bet you they have a fully operational earthbound Voyager copy, and test these things extensively. That probably relieves some of the anxiety.

They do (likely, several). The article notes that they replicated the flipped bit that was causing the issue, and demonstrated that it caused the same symptoms.

I wonder, is it still possible to get involved with the Voyager craft, i.e. secure employment at JPL and get into that team? Voyager 1 & 2 are hands-down the most incredible machines/computers we have ever made, IMHO.

I hear you. I actually managed to wrangle Voyager 1 into a brief nerd-out component of my wedding speech:

"Like me, the space probe Voyager 1 launched in 1977. Unlike me, it is one of the pinnacles of human achievement. It is currently the farthest man-made object from the Earth but it could not have achieved its speed and distance to date without gravity assists from Saturn and Jupiter. When space probes are launched, their path is often planned to get a gravitational slingshot from the most significant entities in our planetary family, to take them further than their launch alone might do. That boost is something invaluable, natural and efficient."

"I don't want to call my parents, (mother) and (father), giant balls of gas, but the boost, the impact they have had on my life, on my sister's life, my brother's life and now the lives of our partners, is huge." And then, later on: "Everybody, wish me luck that in our future together, there are only few occasions where (bride) wishes I was the farthest made-made object from Earth."

You are awesome. I hope your comment gets into Hacker Monthly.

And I thought making a presentation about culture shock with Dr. Spock was good.

Hey, thanks. I also managed to incorporate a bit about companion planting and its similarities to relationships in there without it being too bland or dry.

It was all well received and only trumped by my father who wrote something very original that had people in stitches. I should put both online just in case other grooms are looking for inspiration - almost all of what's out there is pretty basic and there is little worse than a cookie-cutter wedding speech.

Maybe but it sure as heck won't run Ruby.

Frankly, that's part of the beauty.

Sometimes it's worth springing for the ECC memory.

Launched in 1977 - wow.

(See http://en.wikipedia.org/wiki/Voyager_2)

Distance: 8,600,000,000 miles

Radio wave: 186,282 miles per second

Push instructions: 12 hours, 49 minutes, 12 seconds

Flip bit: ?? milliseconds [+]

Wait for sample data: 12 hours, 49 minutes, ?? seconds

Process sample data: ?? milliseconds [+]

Analyze and confirm sample data: ~4 hours

[+] does anyone know this value?

It's a continuous signal so it had to have been exciting to view the stream at the appointed time and see the correct format. Kind of like setting the baud rate and parity on an old VT100 and getting the username prompt.

Congratulations to everyone involved.

It is kind of depressing to think this spacecraft that we launched over 30 years ago is only 1/2 light day away and the next nearest star is over 4 light years away.

More good info here: http://voyager.jpl.nasa.gov/science/index.html

Batteries should be good until 2025, data is sent back at 160 bits/second

500 years ago the only way a human could go faster than, say, 20mph was by jumping off a tall structure. Even 200 years ago nobody knew much about the nature of the Universe outside our own Solar System, nobody knew how atoms worked, nobody knew about fission or fusion, nobody knew how the Sun worked, automobiles, computers, heavier than air flying machines, and spacecraft were as fantastical and unbelievable as any Greek myth, perhaps more so.

Since then humanity has grown by leaps and bounds. Conquered diseases. Conquered limitations of distance. Learned many of the answers to the Universe's deepest questions. We know how stars work to a degree that would astound an astronomer from 1800. We know the age of the Universe to an incredible precision. We've spied on the microwave echo of the big bang. We've sent humans into space repeatedly and landed on a neighboring planetary body. The Universe is a big place, we've only barely started to explore it and understand it, but we shouldn't ignore how far we've come.

Horses and some sailing ships go faster than 20 mph, but yeah.

Also, Voyager wasn’t designed for maximal speed out of the solar system. Even then, if we had wanted to get to Alpha Centauri as fast as possible, we could have done better. Not that the numbers aren’t humbling.

Pinging VOYAGER2 with 20 bytes of data:

Reply from VOYAGER2: bytes=20 time=46152000ms TTL=53

And I thought my latency was bad...

The Voyager probes are absolutely astounding. I thought President Bush had proposed cutting funding to that project as a cost cutting measure. I'm glad that never occured.

Reminds me of Ron Garret's article on Lisp at JPL ...

"Having a read-eval-print loop running on the spacecraft proved invaluable in finding and fixing the problem."

http://www.flownet.com/gat/jpl-lisp.html

Following the chain of articles to http://www.flownet.com/gat/jpl-lisp.html - Lisping at JPL and seeing that essentially lisp was abandoned because of an unreliable inter-process communications app written in C makes me wonder what SpaceX, Boeing etc are using.

It's so easy to anthropomorphize spacecraft like Voyager 2. I get choked up thinking about her bravely leaving the solar system, transmitting data until ~2025, alone in empty space.

XKCD nailed this perfectly when the Spirit rover got stuck and NASA had to call it quits.

http://xkcd.com/695/

There's a story by Vernor Vinge called The Long Shot you might want to look into.

33 years and going .. Congrats to all

Images from voyager : http://ffden-2.phys.uaf.edu/212_fall2003.web.dir/Brian_Herol...

From Wikipedia : "It is possible that one or both craft may have enough RTG energy to last until 2025, but there is only a small probability of this"

http://en.wikipedia.org/wiki/Voyager_program

everyone needs to read about voyager 2. among other amazing things, its engineers designed it so that it would stay functional in various ways as its battery depleted over the decades. if it does not receive a transmission from earth after a given time interval it will go into some kind of autopilot, faithfully beaming data back to earth. third thing i recall: the signal from the probe is about as faint as a cell phone on the other side of the world

In various papers I had to write in college, I'd tend to base them on space probes whenever possible. A great excuse to read journal articles about the Lunar Prospector mission and so on.

I know its off topic, but when you see this dedication to quality, and the aspirations that led the Voyager mission, you cannot but feel saddened and stupefied by the clusterfuck we are witnessing these past 2 months in the gulf of mexico.

{edited}

I don't follow golf so I don't understand what you are talking about. What's the relationship between golf and NASA missions?

I think this this raises an interesting point, why do we look at space with such awe in comparison arguably more complex and certainly more dangerous engineering tasks closer to home?

I agree that the point is interesting.

Part of it is the differentness of the environment (high radiation, low temperature, microgravity, great distance). This necessitates a rethinking of design constraints at a fundamental physical level -- i.e. deriving system constraints from first principles.

Another part of it is the unique nature of a lot of the stuff we send out there -- rovers to Mars, planetary orbiters with exotic cameras, Earth-orbiting satellites with hyperspectral imagers or advanced radars, powerful telescopes with cryogenic optical/infrared cameras.

Here's an example (.mov):

http://marsrover.nasa.gov/gallery/video/movies/mer_ch_edl_Te...

from

http://marsrover.nasa.gov/gallery/video/challenges.html

A building on water doesn't seem that different to the regular world for most of us. When most of what you see is on the surface (aerial shots of oil rig, smoke plumes, oil slick, etc), 95% of the population probably wouldn't realise that the rough stuff is happening over a kilometre under water.

In case you were wondering where the spacecraft gets energy, it's powered by nuclear decay:

http://www.nytimes.com/1989/08/26/us/voyager-s-heartbeat-is-...

Is amazing at the wealth of knowledge these older guys have.

Funny, I had heard some article trying to say that voyager was sending 'alien messages'. Of course it's just one bit swung the wrong way.

What caused the bit to be flipped? Cosmic rays?

Exactly my thought! Remember this one? http://blog.ksplice.com/2010/06/attack-of-the-cosmic-rays/

do they just telnet into it? anyone know the login?

seriously though, how do they go about accessing the system?

First, you need a 70m antenna.

http://deepspace.jpl.nasa.gov/dsn/

http://deepspace.jpl.nasa.gov/dsn/images/realtime/latestS.jp...

And, it seems simple, but is not: you need very precise pointing, and very precise ephemeris (http://ssd.jpl.nasa.gov/?horizons) to get the relative velocity to the s/c. You need velocity so you know what the Doppler-shifted frequency of the transmission will be.

Furthermost SSH ever?

that always works

Reply on news.ycombinator.com