Recorded on February 26, 2019 at G-Unit Studios in Rhode Island!
- RSA Conference 2019 is coming up March 4 – 8 in San Francisco! Go to rsaconference.com/securityweekly-us19 to register now using the discount code 5U9SWFD to receive $100 off a full conference pass! If you are interested in booking an interview or briefing with Security Weekly, please go to securityweekly.com/conferencerequest to submit your request!
- Join us April 1-3, at Disney's Contemporary Resort for InfoSec World 2019 where you can connect and network with like-minded individuals in search of actionable information. Visit https://infosecworld.misti.com/ and use the registration code OS19-SECWEEK for 15% off the Main Conference or World Pass. If you are interested in booking an interview or briefing with Security Weekly, please go to securityweekly.com/conferencerequest to submit your request!
- Registration is now open for the first Security Weekly webcast of 2019! You can register for our "Rise Above Complex Workflows: Practical Ways To Accelerate Incident Response" webcast now by going to securityweekly.com/webcasts.
Topic: Understanding Storage
- Everything you do on a digital device has to sit somewhere. That stuff on the screen you are watching, it's only sort of there. A bunch of pixels which are flickering faster than you can resolve creates a kind of animated page where the screen is constantly being "refreshed" from somewhere else. In fact, the screen is just one instant of an image and there is another one being drawn in the background faster than you see. Basically, it's all just like a flip book where you draw those little animations of the balloon floating away, sadly as you gaze out the window of the classroom humming, Helena to yourself.
- So, where is it if it's not on the screen? Well, it's in memory. It could be on a disk (which can also just be memory). So, what does that mean exactly. Well, there are two basic kinds of memory that we see being used right now. One is called RAM which is more accurately "flash" (not to be confused with the horrible web language, Flash). Flash ram or usually just flash is essentially a chip. A chip is really just an integrated circuit which is a fancy kind of transistor. These things are powered by electricity and have on/off conditions (zeros and ones) which allow them to create groupings we call bytes. Computers don't really care how they get things, so long as they are in binary form and subsequently, the hardware doesn't really matter to the computer just that there are groupings. They could be 8 bits, 7 bits, doesn't matter. The operating system is what determines all that. But the chip just has a lot of switches that can be used.
- That introduces the term "volatile" and as such we also have "non volatile". Volatile storage relies on electrical power to maintain it's state and if the power is lost well, so is whatever is there. Non volatile has some other means to "save" or preserve the state of the array based on some other system. Originally, all flash memory (RAM) was volatile and mostly it still is if it is being used for RAM. That means, if you drop the power, you lose it. This is the fastest form of storage that we see being used since these types of systems are electrical in nature and as such travel at very very high speeds. Non volatile memory is then flash that can be "written to" and as such can be preserved or could even be physical in nature. Now, originally, chips had RAM chips (volatile) and ROM chips (Read only) (non volatile). The ROM was used to store what is now typically called "firmware" and the RAM was used to store current things. When the computer powers up, ROM is used to bootstrap the hardware and create a space on the hardware for RAM to be configured.
- Now, at point X in history, RAM was very expensive per bit and as such as little as possible was used. Because of that, cheaper, physical storage was the norm for most things. This type of storage typically used magnetic locations which could have their polarity set to + or - (1 and 0) and the state could be preserved because the magnetic nature of the storage held up without electricity. So, tapes, floppy disks, zip drive, and even hard drives all used this type of technology to store things long term or things that didn't need to be accessed quickly.
- What we have seen in the last ten years is a convergence on the "flash" model of storage due to the ability to save states in integrated circuits that is rapidly approaching the price per bit of magnetic media. Now, in 1986, I bought a 10MB hard drive that cost about 2000 dollars. Today, well, those prices have fallen. But at same time, chip based flash storage has dropped dramatically in price such that we can use RAM type chips for storage. That means that now, SSD drives which are just chip based, are becoming the standard since they are fast, quiet, low power, and durable. Poor old platter based hard drives are dying out.
- Modern Storage for SSD and other non volatile devices have two main types in use NAND and NOR. NAND is more dense and is subsequently bigger. NOR is faster but costs more per bit. NAND is usually what is being used in SSD drives. It has a finite number of times it can be written to and as such won't last forever (but what does?). This is the basis for "flash drives" as well as "SSD hard drives". All these things usually use NAND chips which allow for the preservation of state even when the power is turned off and as such provide very fast, quiet, and cool, operation compared to the physical hard drives which have power supplies, spindles, platter, read heads, etc. all of which require power and have a higher likelihood of failure.
- So, how does it work? Well, An old EEPROM chip hd the ability to store data and be overwritten but it was an arduous process. We wanted fast, small, quiet, etc. so enter flash. Now, a normal transistor has three connections inside of it: A source, a gate, and an output. Electricity comes in the source and is stopped by the gate so it can't flow on out the output. If you close the gate, no flow, the transistor is "off" or 0. Open the gate, the transistor is "on" or 1. Boom. The problem is, you gotta have power. So, let's use our EE skills and add a floating gate before the regular gate. Now, we can open one gate and close the other and the electricity is trapped by the floating gate inside the transistor. Thus, the on state can be preserved even when the source of power is removed. This creates a flash chip which can then preserve the ones and zeros even without power. Flash drive is born. The floating gate eventually wears out after 10000 or so uses but it really just slows down as it degrades and as such, it may last a LOT longer but maybe not as nice. It's said that somewhere between 10000 and one million cycles it will fail. In my experience that has been longer than I usually had the device.
- We still see traditional platter drives being used for LARGE storage arrays because the number of writes and spins is far longer than SSD typically, but even those devices die eventually after enough spins so the graphs continue to converge in both price and longevity.
- How about DNA. Yep, you can jam data into biological strands too. I read an article that said a dna based drive the size of a teaspoon would store every piece of data in the world and it doesnt' degrade apparently. Wow. Kinda pricey though still.
- More likely, Quantum. Now, we talked about quantum computing recently so I won't get back into that but if we use those Qbits we can vastly expand the number of states that exist at the same time in a much smaller space. That means that the flash of the future may be Qflash which would not only allow for massive storage in a tiny space but could also use things like Quantum Entanglement to create multiple automatically synced copies of the data instantly.
- All this is still just a means to create groupings of zeros and ones that can be read on the fly and used to create cat pictures on your screen so you can get a larf but that's how it works. Now, get out there and invent something better.