EvolvedKiwi

Apologies for the confusion, mixing terminologies up is something im golden at. The 'cache' is indeed a pool defined with SSD's instead of spinning disk for that performance bump with the docker and vm usage i have planned. Thank you for the clarification on that, also thank you @Kilrah for your response also.

I've recently acquired a R730xd LFF and would like some advice on my plan for configuring the array. My original plan was to use 2 x 1TB ssd's for the array but this isnt viable as i've read that Trim is not available due to the parity being mismatched between the drives after the Trim command is run. I have 12 x 3.5in drives that i had planned to use for the pools while using the 2 x 2.5in bays at the back for the ssds for the array with no cache. If i have to mess around using 2 of the 3.5in drives for the array that messes up my pool design. My question(s) is: can i set a USB dive as my array drive with no parity drive and then create a mirrored cache while setting all folders to run off the cache directly with no Move being carried out back to the array? If this is possible, what problems am i likely to encounter with such a configuration? Im unsure if the cache needs to be btrfs or xfs given btrfs supports discard=async mount option while xfs needs to be manually trimmed (i think!). Any help would be appreciated, thank you.

Im seeking guidance on how to go about building out my home lab and would really appreciate help. A local company had decommissioned and was disposing of the following hardware which i've acquired (this equipment was in storage for 5+ years): 1 x IBM x3550 M4 (2 x E5 2620 v2 CPU, 128GB RAM) 2 x IBM x3550 M3 (2 x E5649, 96 GB RAM / E5 2620, 96GB RAM) 1 x IBM DS3524 Express Storage Unit - mix of 146GB/300GB/500GB/1TB SAS drives 1 x IBM EXP3524 Express Expansion Unit - mix of 146GB/300GB/500GB/1TB SAS drives My background is in desktops (custom builds etc), software support and recently becoming a junior devops engineer at my place of work. I have not worked on server grade equipment before. My final goal for this project is: Plex server large storage array local kubernetes cluster local docker windows/linux vms move my pi-hole, synology ds414 and qnap ts412 in with the server look at getting a managed switch (currently just using basic tp-link switches) I've tested all the x3550 servers and am able to access the bios, one of the server's had a bad ServeRAID M1115 raid card but the other 2 appear fine (one is a M1115 and the other a M5014). Im only planning on running the M4 at present until i need/want more compute (and a bigger power bill). I managed to create a driver update package for the M4 and upgrade it to the most recent firmware and am using IMM. I've started the M4 using Unraid as it has native docker support, containerized Plex and a few other apps im interested in and this is where im starting to go sideways. I've been reading about cross-flashing the ServeRAID cards back to their LSI equivalent but have yet to figure out how to get the Express storage unit functioning and accessible through to the server first. The documentation i've read has left me a bit ???, from what i can work out, the express storage is configured with dual ESM and 1GB iSCSI host interface and the M4 has an expansion NIC. Im trying to establish: How to get the DS3524 configured and recognized? - i cant download IBM storage manger v10 as i dont have a service contract Do i need to flash a raid card in the M4 if the sas drives can be converted to jbod off the raid controller? For increase capacity, can i put 4TB samsung evo ssd's in the m4 directly? - i believe yes from my research I've been reading the forums, servethehome guides, wiki.unraid and a few other places but am hoping for some guidance on where to go from here. Thank you.