Tux in Space
Anonymous Coward writes "In between all the bad news about Linux it's nice to see some good like this story about Linux getting used by NASA for satellite control I saw on NewsForge. It is on Linux Today too. This is not some garage dream but a real NASA project run by a real rocket scientist. Does ThinkGeek sell 'Tux in Space' T shirts yet? I want one!" NASA has a page for FlightLinux.
BlueCat looks like a nice embedded development system and probably excels in that area. It would undoubtedly be suitable for a PDA or an appliance. But it isn't a real-time system. And you need that for flight control.
For reference of what we were looking for, our application is a flight control computer for a model aircraft. Among our criteria was "hard real-time" (stringent timing tolerances on OS response to interrupts. Milliseconds matter. Consistent response times matter.announcedannouncedannounced) An unmodified Linux kernel has some real-time features, but only "soft real-time" which isn't good enough for flight software. So we looked at a number of systems including RTLinux, TimeSys and MontaVista which do offer hard real-time extensions to Linux.
We selected MontaVista's real-time scheduler and kernel-preemption patches because they offer hard real-time without losing access to the POSIX interfaces of Linux. TimeSys also fit that description but MontaVista was better documented and more recently updated (including one in LWN this week.)
I'm surprised that the FlightLinux project didn't have hard real-time criteria at least somewhat similar to ours.
Mind you, in my opinion (as an AirMISR user) they would have been better off using a real RTOS...
Nick
-- "It's a sad day for American capitalism when a man can't fly a midget on a kite over Central Park" - Jim Moran
"In between all the bad news about Linux..."
What BAD news?
Co-founder and designer at Music Nearby: http://musicnearby.com
While NASA using Linux is very good, when will they trust it enough to use it as the primary OS for the main computers operating the Space Shuttle and the International Space Station? After all, extreme reliability of the software is very important for such applications, and I'm not sure if Linux with its current 2.4.x kernel has passed NASA's extreme requirements for such applications.
Raymond in Mountain View, CA
You are incorrect. As a spacecraft thermal engineer, I use radiation and linear conduction as the primary method of heat transfer. What you are thinking about is convection, which requires gravity (and air).
You would still have cooling with the presence of air, but not as much. Plus, for a long science mission, your air will eventually leak out. This is the cheap and dirty method used by Russian engineers, but they have to replace their satellites more often.
I don't know the software (or hardware for that matter) causes of heat dissipation, but I typically have different operating and standby modes to design for. I doubt Linux and provide anything different than currently being used. The only possible benefit I can think of is using slower clock speeds to accomplish the same task as other embedded systems, which means less power, which means less headaches for me. As far as APM goes, it can be a benefit, but a little one. Heaters are often controlled to make up the difference in peak power modes and nominal modes or standby modes.
Good luck with the project, I'm sure it will benefit it the industry!
~afniv
"Man könnte froh sein, wenn die Luft so rein wäre wie das Bier"
~afniv
"Man könnte froh sein, wenn die Luft so rein wäre wie das Bier"
Richard von Weizs
I'm just waiting for all the BSD people to come out, proudly displaying the chips on their shoulders, and decrying NASA for having picked such a non-free, unstable OS as Linux when BSD was perfectly available and they wouldn't have been forced into not making a profit on the spacecraft.
Need a Python, C++, Unix, Linux develop
Thus is Microsoft's greatest chance for increasing NT's uptime lost. Given that the typical spacecraft sees sixteen sunrises every 24 hours, MS would have been quite justified in saying that NT was able to stay up for a hundred days without bluescreening.
And if this OS is for use on spacecraft, shouldn't it be called floatlinux? I'll concede that FreeFallLinux probably wouldn't go over too well.
Yes. And given the relative cost of putting an experiment into space, the cost of a commercial RTOS is *nothing*.
And more people know it? Get real. It's not about administering a unix box.. it's about embedded programming, which is rather very different.
The point is that although linux might be cool for some of this stuff, it's not a gift from god for embedded work. And it's certainly not 'more able to operate in the harshness of space'
especially because in space, no one can hear your screen...
- Microsoft to unveil
.NET software for non-Microsoft platforms
- Defense Department plans 512-processor Linux cluster
- A 'free' interview with Linus Torvalds by CNN
- Linux looms on the mainframe horizon
Just wondering why all the doom and gloom. It's a great time to be a Linux-head!Or, were you refering to the fact that Linux companies which were riding the tech wave without real business plans are getting hit just like everyone else? Companies who were relying on the stock market to make them profitable are going by the wayside, but I think the Linux industry is here to stay. Companies like IBM and Motorola will continue to see huge returns from their Linux investments. Countries like Mexico will continue to use it. But, most of all, the companies that did have business plans and did plan on becoming profitable will have a fighting chance just like any startup businesses.
Disclaimer: I'm a VA/Linux stockholder, so I may be biased on some of this.
I don't want to spoil all the fun but I can't help but laugh when I see someone noticing a story about how Linux is used in some space related case and immediately everyone is talking about how Microsoft is going to die and how this is a sign of how superrior open source is.
Guys.. There have been laptops on just about every shuttle flight so far, for use in various tasks. The laptops have usually been IBM thinkpads and the OS has usually been NT. The fact that someone is using Linux is nice, but it doesn't prove shit - one way or the other. All the other thousands of satellites in space use something else, you know...
It's the same when ONE city in ONE country (Mexico) decides to install Linux on both of their computers and it deserves a news here and a discussion with several hundreds of enthusiastic posts about world domination. Meanwhile, I'm sure Microsoft had hundreds of installations of Windows 9x / NT / 2k elsewhere in the world.
Now I have nothing against Linux - I develop software for it at work as a matter of fact - but I'd like to see some realism and balance in the news here. I don't know about the rest of you but I'd like to read "news for nerds, stuff that matters" - not just braindead pro Linux propaganda that isn't anchored in reality.
From the linux in space page:
Software portability is of vital importance for upgrades and applications enhancement. Portability among UNIX flavors can be done quickly, preserving expandability and keeping manpower costs down. This is not true for other non-UNIX operating systems.
A 'certain other company' ahem.. has spent years ensuring incompatibility between products to tie you into theirs, then make you upgrade regularly. It's very nice to see this policy backfiring on them in a high-visibility market space I'm sure they'd like to 'own' too.
EZ
"Oops, I always forget the purpose of competition is to divide people into winners and losers." - Hobbes
_ _ _
I was working on a flat tax proposal and I accidentally proved there's no god.
let me answer a few question...
PAZware "Dont they usaly write an independant os for satilites and telescopes and such?"
no, normally Sattelites use Propritary, COTS Embedded system like VxWorks
mindstrm "To claim it's 'more able to deal with the harsh radiation of space becasue it runs cooler because its halts the processor for brief times' is rediculous."
we never said that, we have not done any testing but because of the cooling APM features in the kernel it might be better. The problem is, in space heat does not radiate away from components. in order for heat to radiate, it must have air to radiate with, no air and you need to design a cooling system to keep the CPU from overheating. (like when you overclock, and the air circulation is not enough to keep the CPU cool, guess what, we have the same problem, we're not overclocking, but there is no circulation, so we need to find other ways to keep systems cool. the APM Idle functions in only one little but that might help reduce the need and expense for other cooling.
I_redwolf "Linux has been into space before this. Specifically in 96-97"
yes, Debian has been known to fly on a laptop on the shuttle, BUT a Linux system has never been in control of crucial systems, like that of Command and Control, that is one of our goals
papskier "Now we've got astronauts' lives dependent on linux."
if we weren't currently working on flying unmanned sattelite missions. currently not yet, there are no manned missions using Linux in crucial areas. but that may very well change, and I'm sure when we send men to Mars, and the Computers running those systems are based around Linux 4.2.19 you will be the first to know
RayChuang "While NASA using Linux is very good, when will they trust it enough to use it as the primary OS for the main computers operating the Space Shuttle and the International Space Station?"
that will probably not happen, at least not for a while. Both the Shuttle and ISS are in deployment, meaning that the systems onboars have gone through years of development and testing. You will not see any major change of OS' in most existing systems, (imaging if you spent five years and millions of dollars developing an in house application that runs perfect on Solaris, would you change to Linux or BSD at the drop of a hat? no.) However, you may start seeing Linux in use on new systems.
Sarin "I hope they will give us the sourcecode of the programs they run on that machine, but I don't bet on that. Perhaps we will find out they actually use it to run a slightly modified version of lm-sensors!"
of course! this is Linux, Linux is under the GPL, nothing is available for download yet. currently what we have running is based on Lynuxworks BlueCat, and is really a simple "Hello World" (it's reall not very impressive yet, but we are taking baby steps). The initial work has all been done in house with some assistance from the FlightSoftware and IP in Space groups, once we make things publicaly available (and it will be) source will be included for everything that we work on and can Open up (there are somethings that we will probably not be able to open due to NDA issues, but they will most likely be very specific to the Univ. Surrey Sat (http://www.sstl.co.uk/missions/mn_uosat_12.html ) we will be using for our tests. which is a 386EX) but what we can, we will have available under the GPL
ln_sensors? no, we currently are not using that, although we don't know what we may need in the future.
"Why Linux"
If Star Trek has taught us anything, our satellite technologies get absorbed into new, super technology-thingies that then want to destroy us because we are not the prefection they now seek, or they are looking for their creator.
Will these resulting combinations honor the GPL and release their source? Can the FSF use lawyers to get source from V'ger or NOMAD?
If it was said on slashdot, it MUST be true!
And to think, other bird's wings will be useless!
:)
- Hubert
We have a theory, not substantiated yet, that Linux runs cooler than other operating systems
I always thought that Linux was cool, but that was only metaphorically !Only a computer geek will get exited over something like this!
Now sep back for a second and imagin yourself a normal person that know nothing about computers, much less ever heard about Linux:
Geek: "NASA is running Linux!!!"
You: "Uhh What's Linux?"
Geek: "It's a operating system."
You: Huh?
Geek: "You know Windows, well but better!"
You: "Yhm.... ok." (dork)
hmm... for fun I enjoy launching DDoS attacks against 127.87.42.5
Didn't Beowulf clustering originate at NASA? yes.
Didn't that HAM satellite that was mentioned here at slashdot multiple times run on Debian? AFAIK.
Space related sciences and Linux in combination are nothing new.
However, what is interesting here is that NASA, who tests everything down to the last little nit, has deemed linux as worthy for controlling one of their satellites. Their testing is done more rigorously than anyone.
Especially since they don't have Feynman to call when it goes wrong.
A host is a host from coast to coast, but no one uses a host that's close
A lot of on-orbit satellites CPUs are running state machines, not really an OS...most thoughts about CPUs in space have been considering Lynx or Vx Works.
No one said anything about using Linux in mission critical Space Shuttle applications...the article specifically mentions satellites only.
If there was ever a malfunction, as there might be. How much quicker would it be to send out an SOS to linux hackers asking for a bug fix. Be a lot quicker for a few thousand developers around the world trying to fix the bug than a few engineers in NASA. You could become a world hero saving a few lives by fixing the unknown bug.
-- Cheer, Cheer, The Red and the White.