Best dataset-layout for apps between SSD and HDD pools (Plex and Immich)

Hello

So im setting up Plex and Immich on TrueNAS for the first time, and would like to check with more experienced people if im going a good or bad route.
I have one SSD-pool and one HDD-pool. Im using SSD-pool for apps-install in TrueNAS.
But to ensure best performance for apps running, I would have most of the data on HDD-pool as there is “limited” space on SSD-pool.

Plex
So for Plex, I have tried to figure out what might use lots of space, and what wont.
Is the following a good layout considering speed/storage-usage for the Plex-app?

Immich
When it comes to Immich, for now i have around 100GB of photos that will be added to the app. However, i will need to plan ahead for up to 1TB of photos and videos (when/if i get family-members onboard).
Its my first time using Immich, so i have lots to learn here before “going live”…
For Immich, i havent created any datasets yet, but would something like this be good?
SSD: thumbs, profile
HDD: library , pgBackup , pgData , uploads , and video

System

System OS:
TrueNAS Scale 24.10.2

System
SuperMicro X10DRI-T4+
2x Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz
256GB ECC RAM

HBA IT-mode
1x LSI 9300-16i
1x HP220/LSI 9205-8i

Cabinet:
InterTech 4U-4424
6 rows of 4-wide drive-slots
Each row has its own SAS/SATA backplane, one SAS-cable to each row from the HBA’s.

PSU:
Corsair HX1000i 1000Watt

Pools

Boot
Mirror:
2x 480GB HPE MK000480GWUGF SSD

Storage:
SSD_Pool
1 x RAIDZ2 | 6 wide | 894.25 GiB
6x 960GB HPE-branded MK000960GWSSD

BygdePool
3 x RAIDZ2 | 6 wide | 12.73 TiB
12x 14TB drives HPE-branded MB014000GWUDA
6x 14TB drives Seagate ST14000NM001G

Thanks for any help and suggestions

Plex :

Put app on SSD
Videos go to HDD

Immich :
DB goes to SSD
Photos go to HDD

Keep in mind, you don’t want to spin HDDs for no reason just because Plex wants to write to log file.

There is absolutely nothing special in Immich. Its pretty much Postgres DB with metadata + thumbnail cache and your original photos.

You build the database once and after that its simple rescans to pick up changes.