Hard Drive Size Settings

I have a 3Tb hard drive for the storj node.

Configure the hard disk with 2.4TB. To leave 10% free space. The node says it is almost full with 2.29TB.

Windows says the disk has used 2.3Tb used or 2.09TB. And it has 698Gb or 650GB of available capacity.

Is there a confusion with bytes (b) and bits (B)?
Can i increase the capacity to 2.7TB in the node configuration? For my node with 3Tb disk.

Storj uses TB, Windows uses TiB.
Take the 3TB and subtract 10%; 2.7TB should be good.

You have left 20% free, not 10%. I see no problem increasing it to 2.7TB.

The confusion is not between bits and byte, but between Terabytes (3 trillion bytes) and Tebibytes (TiB, which is a power of 2). M$ use an incorrect abbreviation here. The capacity of your 3 TB drive = 2.72 TiB.

Oops, @thej beat me to it.

2 Likes

Thanks for the answers.

I have changed the capacity to 2.7TB

actually TiB is usually a bandwidth definition… because of the mathematics of how additional lanes increase bandwidth. while hdd capacity increases based on decimal addition…

so really when one is talking transfer speeds it should be in TiB and when talking of storage it should be in TB … atleast from an engineering / architect point of view…

ofc that is very confusing… as data transfer would take up a different amount of space on the drive when stored rather than what it was when being transferred…

so yeah doesn’t really make sense… aside from if one is trying to calculate the speed of a certain port based on how many lanes it has… then all of a sudden the TiB definition can help you to calculate port and bus speeds in your head, something which much more counter intuitive when using decimal numbers…

did a deep dive into this a while back… so while it’s a good idea there is a standard… however both methods have their merit

and you know what… i’ve been using windows for decades and i’ve never had to deal with that issue before going over the linux… and all of a sudden the confusion is complete… i can barely put a number on how much space i actually have…

microsoft haven’t changed this because this was how it was done for decades, and it would be a very confusing change…

if i was talking with other people using windows mainly, then there would be zero confusion when i said 1TB :smiley: and its been like that since i started using it, no matter what people want to call it today…

not that i want to defend MS because they have gotten to big for their customers to benefit from it… now its like a big tree of the forest with rotting roots… their products are antiquated in most aspects and slightly more stable… and easier to use, but gives less options…

Transfer speed is measured in b/s or B/s, so in your example it should be TiB/s.

it should make sense… ofc… i was just trying to explain why it is like it is…

if you think of a bus like say a pcie bus, then for everytime you add a lane to the bus, the bandwidth goes up by a factor of two… thus TiB makes good sense when calculating transfer speed… atleast on a fundamental level…

while storage… if you add a plater then you simply use addition, which is easiest for us humans to do in decimal.

what people what to use for showing the speeds can be completely arbitrary from these fundamentals… if it was up to me… we should use a 3rd definition that looks nothing like the others… just to avoid the confusion cause by the switch from TB which is TiB becoming TiB
and SI becoming TB…

sure when it’s all over it will make more sense… but really in some cases it will also be quite the mess… because try rounding numbers that are a factor of two into being nice looking decimal numbers…

how do you write 4096 or 8192… ofc the simple answer is just to cut it down and call it 4GB… but it really isn’t… no matter which way one changes it, then binary doesn’t convert well into decimal numbers… just like duodecimal numbers doesn’t translate easily into decimal…

how would you write or count a clock in decimal numbers… sure you can write it with decimal numbers… but you end up doing all kinds of weird stuff…

like each day and night cycle is 1 = 12 and 2x is 24 hours , then each hour is split into 5 parts of 12 minutes, giving a total of 60minutes in an hour…

ofc the setup kinda drifted a bit over time… but you get the idea… and it simply cannot be easily translated into decimal… no matter how much one wants to do so…

the fundamental behavior of the numbers and even numbers will not align into something sensible… so tho i think it is kinda arbitrary what is used… then i don’t see how it’s even possible to change it like people seem to want to do… is simply not how computers work…