back
25 comments
The lab I work in at NASA Goddard is doing the detector characterization for WFIRST. There are 18, 4k × 4K detectors in the WFIRST array, for a resolution of about 300 megapixels, as mentioned in the article. The detectors are arranged in an unusual pattern, but I don't know what drove that design. I am, at this moment, working on a new fileserver that will be used for WFIRST detector test data. It will start off with a RAID array of just under 1PB and will likely be expanded as we proceed with testing.

I'm hoping this thing launches before I retire, but I'm not betting on it. JWST keeps eating our budget.

Quietly hoping you’re about 64 years old.
Just 61-1/2! But I may not wait until my official retirement age of 66-1/2, which I'll reach at the end of 2023. I don't seriously believe that WFIRST will launch before then anyway. I'm not even sure our other major project, Euclid (an ESA telescope) will launch before I'm retired, and it's much further along.
Sometimes I really love the 21st century. "Yeah, we're starting with a petabyte... it's not really enough, it won't last that long, but it'll get us started, at least. I've already started on the purchase orders for the next three petabytes...."
Any infos about the RAID itself? (e.g. disks, raid type, software managing the raid, OS it is running on, client protocols being used, which decisions/priorities resulted in the current setup, etc...)

I love such details :P

It's nothing really exotic. It's a "high density" storage system from RAID, Inc., with 84 drives in a 5U form factor. We got 12TB drives, so raw capacity is 1008TB. It's connected to a Dell R640 server running CentOS 7. I'll configure it with RAID 6, with a some hot spares. I'll also buy a few cold spares. The file systems on it will be shared via NFS to various client systems. We have a rack of compute servers for data analysis which are connected to the file servers through a 10Gbit switch.
Just wondering - why RAID6 and not ZFS?
I don't have a lot of experience with ZFS and I've been reluctant to use it on Linux, due to the licensing issues. But I am interested in ZFS and will probably look into it at some point. I'd probably need a different storage solution, one that presents all the drives individually to the host, since that's what ZFS wants.
Cool project! ZoL is rock solid and it's easy to setup. You could switch it over in a weekend. The standard HBA last time I bought one was the M1015 flashed into IT mode. ZFS data integrity features alone are worth it for that much data. If you mention what you are up to on freenode#zfsonlinux people will jump at the opportunity to help.
Any chance at bringing someone new in to take over the brunt of the design and implementation? It sounds like a natural point to transition.
Isn't RAID6 iffy re bit rot at PB levels?
…and the addition of one of the most sensitive coronagraphs ever made…

I find it interesting that one of the major improvements in this light gathering instrument is a device to physically block light from getting into the instrument.

For Earthlings viewing the Sun a chronograph is a disk on a stick to block the Sun's surface so you can see the corona.

In a space based telescope blocking out a star but not any planets which might be orbiting it becomes complicated. They have intricate computer generated masks designed to block the starlight and the interference patterns created by the edge of the primary mask implemented as mirrors on black silicon.

I wasn't able to find a published paper on the design, there are several available as abstracts but not viewable by me. This presentation from Caltech covers a lot of it and is probably accessible to people in the field. http://kiss.caltech.edu/special_events/JPL_MPIA/presentation...

My takeaway was that compared to the Hubble, they should be able to block 100 times more starlight relative to the planets and should be able to directly see planets illuminated by their star.

>"found that they had a couple huge telescopes just sitting around in a black-site warehouse somewhere"

i don't know what is the worse implication, the amount of bureaucracy capable of losing a space telescope or the other technology the NRO and other gov't agencies have hidden from the public view

That's a pretty uncharitable mis-reading of what happened. When a spacecraft is being manufactured there is often an additional, identical space-ready backup article built alongside the primary vehicle. This is so that accident, error, or other mishap don't disrupt the schedule for the craft being developed. If some component tests bad, even as late as when the satellite is being loaded on the launch vehicle, there are space-ready spares, for everything, available. If the vehicle blows up on launch you can assign the backup article to an available launcher. Also, once the satellite is launched they have an identical test article available for troubleshooting and other ground-based testing.

So, what do you do after your mission is a success and its next-gen successor is being built? You surplus the backup articles. That's where the gift articles NASA come in.

None of this is incompetence. It is, in fact, prudent and fiscally conservative program management.

They didn't say they were lost. Maybe they have even better ones now and these are obsolete in the spy business.
This is basically the case here. NRO has newer stuff that is better. Also, they’ve switched to more SAR instead of visual spectrum stuff.
“Despite having the same size mirror as the surprisingly reliable Hubble Space Telescope, clocking in at 2.4 meters across, this puppy will pack a punch with a gigantic 300 megapixel camera, enabling it to snap a single image with an area a hundred times greater than the Hubble.”

Sounds like we need to retrofit the camera on the Hubble.

As Hubble nears its EOL, maybe some crazy billionaire will come up with a plan to keep it around for another 25 years.

The problem with Hubble is not so much funding, but that the hardware is dying, especially the gyroscopes. Without the space shuttle, no one can just fly out to service it. It would be easier to launch a replacement telescope.
Yes, everyone already knows that!

That's where the crazy billionaire comes in. He doesn't tell us why it can't be done, he figures out how to do it.

electric car company... rockets that land and are reused...

I imagine that you could do it with advanced robotics. Could happen over the next decade.

Or one just builds a new telescope and puts it onto a F9.
A new Hubble costs a couple of billion? James Webb is over $10 billion.

A few small robotic drones could hitch a ride on a regular launch. Mission cost of $10 million?

We simply need to wait for robots to develop and hope it happens before Hubble disintegrates.

Can't one use one of manned spacecrafts being developed in USA to get to Hubble? Soyuz is out for other reasons.
Get to Hubble - possibly. But they lack the ability to conduct the repairs, like they are missing an air lock for activities outside of the vehicle.
It’s nice to see that the donated NRO satellite buses found a use i remember when they were initially donated everyone cried that NASA didn’t had the budget to make any use of them.

I wonder what else do they have lying around, since it’s clear that what they have now is more advanced than what they’ve donated.

I also wonder with the advancement in SAR technology how much longer would optical recon sattlites would even be used for because as long as the NRO still uses optical sattlites we have some sort of economy of scale that can be leveraged for cheaper space based telescopes.

I was curious if this would beat James Webb to space, but this is targeted for middle of the 2020s while NASA appears to be pushing for James Webb to get up to space in the next couple years. Fingers crossed!

> In testimony before Congress in July 2018, NASA administrator Jim Bridenstine proposed slowing down the development of WFIRST in order to accommodate a cost increase in the James Webb Space Telescope, which would result in decreased funding for WFIRST in 2020/2021.

Source: https://en.wikipedia.org/wiki/Wide_Field_Infrared_Survey_Tel...