CODEWITHSUNDEEP
linuxuuidpartition
thisguyonceagain
thisguyonceagain

Reputation: 9

When would you use uuidgen in a live environment?

I came across uuidgen from watching a video to study for the redhat 8 exam but I had a question about it's usefulness and did not find any other thread nor did the manpage mention it. So I understand that each device has an UUID and the UUID can be used for multiple purposes - the purpose in the video was creating new mountpoints and the UUID was used to associate a new partition (/dev/sdb3) with a new mountpoint. Here they pointed out that you can generate a new UUID for that device using uuidgen.

My question is why/when would you use uuidgen in a production environment with live servers to relabel a partition? What would be the point of relableing the UUID for a currently existing mountpoint? Is there a sort of attack that target UUID of a system? Or is the sole purpose for uuidgen just used to create random UUIDs for others things like web links?

Thanks

Upvotes: 0

Views: 125

Answers (1)

mmirate
mmirate

Reputation: 722

Say you have a system with several disks, one partition each, and you need to play "Musical Data" with some of them.

If you start by copying, at block level via e.g. dd, the entire GPT-partitioned disk, then you will have, as a result, a duplicate UUID. This is fine if one of the duplicates is going to be blown away before the next time the OS needs to mount one of them. If, for whatever reason, this can not be ensured, then whichever copy you don't want the OS to pick up anymore, needs a new UUID. Enter uuidgen.

I'm assuming you're talking about GPT partition UUIDs, which are stored all together in each GPT that contains the identified partitions; if, instead, you're talking about filesystem UUIDs, which are stored inside the metadata for that filesystem and thus are copied whenever dd'ing that filesystem, then the above scenario still holds and more scenarios become plausible.

Upvotes: 1

Related Questions