Lunar Gateway networking issues?

[darthgently]

I'm a bit dumbfounded that this could be happening in this day and age. 

Maybe (hopefully?) I'm missing something but a team of above average Cisco or Network+ certified engineers eat problems like "bad network chip" for breakfast.  Toss the boxes with the bad chip?  A depot-level chip replacement ?  I assume it is likely complicated by a bespoke space network protocol and chip but there are manuals the Geek Squad A-team could refer to, right?

...The report also finds that the Gateway program is running into some pretty serious technical difficulties. One involves a defective network chip that facilitates communication throughout the lunar space station. Its failure could cause myriad problems onboard the Gateway.

"For example, these defects could lead the flight computers to unexpectedly restart," the report states. "If the network is not functioning properly, it could result in loss of control of the Gateway. Program officials are also concerned that they might identify more defects with the communication network, based on the number found already."...

https://arstechnica.com/space/2024/08/nasas-lunar-gateway-has-a-big-visiting-vehicles-problem/

 

[monophonic]

5 hours ago, darthgently said:

I'm a bit dumbfounded that this could be happening in this day and age. 

Maybe (hopefully?) I'm missing something but a team of above average Cisco or Network+ certified engineers eat problems like "bad network chip" for breakfast.  Toss the boxes with the bad chip?  A depot-level chip replacement ?  I assume it is likely complicated by a bespoke space network protocol and chip but there are manuals the Geek Squad A-team could refer to, right?

...The report also finds that the Gateway program is running into some pretty serious technical difficulties. One involves a defective network chip that facilitates communication throughout the lunar space station. Its failure could cause myriad problems onboard the Gateway.

"For example, these defects could lead the flight computers to unexpectedly restart," the report states. "If the network is not functioning properly, it could result in loss of control of the Gateway. Program officials are also concerned that they might identify more defects with the communication network, based on the number found already."...

https://arstechnica.com/space/2024/08/nasas-lunar-gateway-has-a-big-visiting-vehicles-problem/

 

Those engineers have access to a myriad of chips that take full advantage of the benefit of Earth's magnetic field protecting them from the radiation of space. Exposing them to deep space conditions would quickly turn them from "network chips" to "random number generators with extra steps." When you combine the requirements of radiation hardening and certified for space, you are left with a very short list of potential chips. It is very possible that only one chip can provide those and the required network performance.

Now, replacing failed chips is one thing. It is a lot harder problem, if said defect is in the design of the chip. In that case replacing chips does nothing. There are only two ways around that. You could use a different chip, assuming one is available. That will require re-designing, manufacturing all the surrounding electronics, which takes time and money. Or you can try and work around the defect in software, to prevent the defect from triggering. That also takes time and money.

Even if it is individual chips failing and not a design defect, there must be time and money spent to understand why they failed and if that could happen again during the operational lifetime of the station. So, no way out of this that does not require time and money.

[Superfluous J]

I'm not sure Best Buy has the budget to send up someone from Geek Squad to replace the chip every time it dies.

And you can't keep the chips on the station (except in a lead-lined box I suppose?) because they'll degrade at the same rate. It'd be like trying to store firewood in the fireplace.

[darthgently]

4 hours ago, monophonic said:

Those engineers have access to a myriad of chips that take full advantage of the benefit of Earth's magnetic field protecting them from the radiation of space. Exposing them to deep space conditions would quickly turn them from "network chips" to "random number generators with extra steps." When you combine the requirements of radiation hardening and certified for space, you are left with a very short list of potential chips. It is very possible that only one chip can provide those and the required network performance.

Now, replacing failed chips is one thing. It is a lot harder problem, if said defect is in the design of the chip. In that case replacing chips does nothing. There are only two ways around that. You could use a different chip, assuming one is available. That will require re-designing, manufacturing all the surrounding electronics, which takes time and money. Or you can try and work around the defect in software, to prevent the defect from triggering. That also takes time and money.

Even if it is individual chips failing and not a design defect, there must be time and money spent to understand why they failed and if that could happen again during the operational lifetime of the station. So, no way out of this that does not require time and money.

I get that, but I suspect the chip isn't simply a proven logic design that has been hardened.  I suspect an overly-bespoke design.  A reinvented wheel that appears it may be a bit square. Which annoys if true

Edited August 2 by darthgently

[Nuke]

this is why computers like mother from alien still make sense in a future setting. you really do not need that much cpu power for all the space nav stuff, the agc managed just fine and it was slow.

you do however need exotic systems for scientific data acquisition. though on the iss it is cost effective to send up new laptops periodically.

[domenprevc]

I recently tackled some networking issues with the Lunar Gateway, and I’m thrilled with the results! Initially, I faced frequent disconnects and slow data transfers, but after troubleshooting, I resolved the problem by adjusting the signal frequencies and upgrading my network hardware. Now, my connection is stable, and the data flow is smooth. If anyone else has faced similar issues, I'd love to hear how you solved them. 

[Jacke]

On 8/2/2024 at 9:35 AM, Superfluous J said:

I'm not sure Best Buy has the budget to send up someone from Geek Squad to replace the chip every time it dies.

And you can't keep the chips on the station (except in a lead-lined box I suppose?) because they'll degrade at the same rate. It'd be like trying to store firewood in the fireplace.

Most electronic components tend to be more tolerant of radiation than people.  Though the people have some biological processes to deal with some of the damage, so there's likely some low radiation environments that would be worse for some electronics than the crew.  As for keeping spares in a lead-lined box, it needs to be a complex protection (same for any "storm cellar" for the crew) due to some shielding causing secondary radiation that has a worse absorption than the original primary radiation.

There are also low-level electronics designs that stand radiation better than others.  So I'd say use more radiation robust electronics rather going cheap off-the-shelf (which is a good principle overall for space components).  Also make these components easy to replace and set up a standard replacement schedule.