Disk Drives
When I was an operator we had 2311 and 2314 disks. When I was a sysprog we had 3330 disks.
These were kind of two part. You had the drive, motor to spin the disk, heads to read/write, and the stuff to support it. You also had the disk pack. The disk pack came off and you generally had more disk packs than drives and you would change packs to run different jobs.
The drives were 2311, etc. The packs were a different number. I have no idea what that number was. We always referred to the drive and the pack by the drive type.
The packs were abound 14″ in diameter and made of a bunch, 5-10, of platters. Each platter was a flat donut with a hub in the center. the hubs were taller than the platters and they joined the platters into a pack. A user could not take apart, and have it live, the disk pack.
To store the pack off the drive you took a clear plastic cover. It had a handle in the middle that could rotate. You opened the drive. Put the plastic cover in. When you twisted the handle a little bit in the cover would screw into the pack. When you screwed in the cover it released the pack from the drive. You could lift the pack off the drive. To put a different pack on you just did the reverse.
Now you took a bottom for the case, made of black plastic, and clipped it to the bottom of the disk pack. This gave you a sort of sealed container with a disk pack inside. The top had the handel, the disk pack was attached to this, the bottom was attached to the bottom of the disk pack and kine of sealed the whole thing.
When you hit Start the pack would spin up. When it was spinning little brushes would come out and sweep andy big stuff off the platters. The brushes pivoted on the edge of the drive so they came out along the edge and swept on an arc to the center. After the brushing the heads would come out and it was ready to read/write. The heads flew over the surface, not touching, but they were nowhere as close as a modern nead.
You had a cabinet with slide out shelves that you would put the disk packs you were not using on.
You might have a disk with the employee records that would only be mounted when you ran the payroll application. There could be another pack that held vendor information. You could mount the employee pack, run payroll, replace the employee pack with the vendor pack and run accounts payable.
When I was a sysprog we also had 3340 drives. These took the great step forward of having the heads along with the platters.. The case for them was a big piece of plastic, with a handle on top, that had a sliding door on the back. Put one of these on the drive and the back door opened and a thing grabbed the heads.
Data Cell
Early on there was a mythical device, the data cell. I never saw one, I only heard stories about them.
When you addressed a disk it had a CCHHRR, Cylineer, Head, Record, to get the gecord you want. One thing a lot of people forget is there was a BB, Bin, on the front. This was a thing in the center of a ring of bins. Each bin held a bunch of strips of magnetic stuff. They were between the floppiness of a piece of tape and a piece of metal. When this thing got a read/write command it would take the BB part of the address and move the circle of bins until the right bin was there. I never said this thing was quick. Next it opened the bin and some fingers came down and split the strips apart to leave the one, CC, you wanted. It grabbed the strip, pulled it out, and wrapped it around a drum, and spun it. Now it used the head, HH, for the track you wanted. On the track it got the record, RR, you wanted. When it was done reading/writing it stuck the strip back and got the next one.
I have no idea if this is real. I am afraid to go to the IBM site and find out it was a fever dream. The address of data on a disk was really BBCCHHRR and the BB part was ignored for disks. I did go to the IBM site, The 2321 datacell is real. Not quite how I imagined it, but real. It stored 400 Meg, far more than a 2311, but far less than the memory on my phone.
This sounds like a Rube Goldberg thing. The story told of it not spreading the strips properly and shoving a strip back in on top of another one.
There was some really weird stuff in the early days. the first dids drive I worked with leaked oil and the CE would put card punches in to soak up the oil that leaked. Early memory, before me, was a pool of mercury that would have little waves represent 1 0 and they would move down the thing, hit the end, reflect back, and be read. They also used CRTs. Light a spot up and it would glow for a bit. The machine I worked on used little doughnuts of magnetic material with X and Y addressing done by a grid of wires and the sense a third wire going through all of them. The machine I worked with had around 2 and a quarter million of them that some lady spent days string up on the 3 wires in each. If they had this strange stuff, a datacell is not that much stranger.