I have been trying to understand Object Storage vs. the other two. So only use block storage RAW like that if you are using a database because of the retrieval and the partial updates. But File Storage and Object Storage sounds literally like the same thing to the end user. It does sound like it is a file share in the cloud that is scalable just like if you used your local file server and kept expanding storage on it, but it has better indexing. It only seems to actually make a difference if you are making web applications or applications that interact with the storage because then you don't necessarly have to worry about the location, only the ID. So you can say that song.mp3=1Dlkhd-jfle4i9n-oihafenkl-nan and then all you have to do to have anyone play song.mp3 would be to have the application pull that ID no matter where the file sits unlike traditional file storage which would require you to hard code file.mp3=\\servershare\folder\folder\folder\ to access. And if you scaled that out then that \\servershare may change over time or it may move folders etc. Seems like W11 should have added UIDs to files so you could basically have the same on your local file system. ...or am I missing something? I get it that the "back end" may be different in how it is stored but I'm sure it's just ZFS or some implementation of that.
So: #1. Filesystem level abstraction of block storage #2. Actual raw block storage (either a physical or logical volume of raw bytes) #3 Pure marketing wank.
I have been trying to understand Object Storage vs. the other two. So only use block storage RAW like that if you are using a database because of the retrieval and the partial updates. But File Storage and Object Storage sounds literally like the same thing to the end user. It does sound like it is a file share in the cloud that is scalable just like if you used your local file server and kept expanding storage on it, but it has better indexing.
It only seems to actually make a difference if you are making web applications or applications that interact with the storage because then you don't necessarly have to worry about the location, only the ID. So you can say that song.mp3=1Dlkhd-jfle4i9n-oihafenkl-nan and then all you have to do to have anyone play song.mp3 would be to have the application pull that ID no matter where the file sits unlike traditional file storage which would require you to hard code file.mp3=\\servershare\folder\folder\folder\ to access. And if you scaled that out then that \\servershare may change over time or it may move folders etc. Seems like W11 should have added UIDs to files so you could basically have the same on your local file system.
...or am I missing something? I get it that the "back end" may be different in how it is stored but I'm sure it's just ZFS or some implementation of that.
So:
#1. Filesystem level abstraction of block storage
#2. Actual raw block storage (either a physical or logical volume of raw bytes)
#3 Pure marketing wank.