What is Compaction algorithm in StorageD?

[aminmaghsodi]

What is Compaction algorithm in StorageD?
I don't see any information about the rocksdb compaction algorithm which is run on Storage In Nebula documentation.

[wey-gu]

Dear @aminmaghsodi ,

it's underlying the rocksdb/lsm-tree compaction :D

ref:

• https://docs.nebula-graph.io/3.3.0/8.service-tuning/compaction/
• https://github.com/facebook/rocksdb/wiki/Compaction

[aminmaghsodi]

As always, Thanks @wey-gu
I suggest to explain minor and major compaction algorithms in Compaction part of documentation. For example let's define in documentation that Nebula's minor compaction (Which is run in background) is X algorithm and when we run SUBMIT COMPACT, it is going to run Y algorithm. Is it a good idea?

Regards.

[wey-gu]

Dear @aminmaghsodi , you are always most welcome!

This is a good idea, we could provide more information on how to correlate different compactions to rocksdb compactions, to enable advanced users like you to know the needed context easier.

Are you interested in PR on this part?

Thanks!
BR// Wey

[aminmaghsodi]

Yes of course I'd like to do it. :)