OSI layers are layers of server software or server hardware?

The OSI layers refer to the seven layer model used for end to end communications. At the lowest layer (Physical Layer), it refers to the physical signal (0's and 1's). At the highest layer (Application Layer), it refers to the application that the end user sees. Here is a good way to remember the layers from the top down - All People Share Their Nasty Delivered Pizza. Here is how they lay out:

A – Application (application request)
P – Presentation (how the data is presented to the application)
S – Session (how the request is transmitted creates session for transport)
T – Transport (Transport Characteristics – what IP port is used, where it is going, etc.)
N – Network (adds information such as IP addresses)
D – Data Link (adds information such as MAC address)
P – Physical (formats the transmission into what the network hardware supports)

This is a simplified version, but when the application requests a file from a server, it sends that request to the next layer down. Each layer going down to the physical layer adds information such as what IP port it will use, what type of session and transport it will use (connection oriented or connectionless), IP address of requesting station and server, all the way down to the zeros and ones that move across the cabling. When the packet comes back, each layer strips off what it added to the packet and moves it back up the stack until the application gets the data it needs at the end.