Hi everyone, I am planning my storage setup and would like to hear your advices/ideas/suggestions/criticisms.
The use case: AI model storage (diffusion, LLMs maybe other models in the future), local code repo, some scraping and personal data potentially used for training.
Requirements:
- 24 TB expendable to 48 TB or more in the future.
- no need for it to be online all the time (happy to turn it on/off when downloading/retrieving data).
- some form of backup/redundancy in case of drive failure/lost sectors.
- cost effective.
Initial plan:
- second hand server rack (Dell 730?) with 8x3.5 bay
- TrueNAS SCALE in RAIDZ2
- 4x8TB HDD 5400 RPM (is is worth getting the MTBF 1M or 2M?)
Current cost estimate: server ~300 USD, drives 4*150 = 600 USD = ~900 USD (note that I do not live in the US, but I translated the prices I can reach to USD for convenience)
Any feedback?