The 2008/2108/3008/3108 and related HBA chips and cardsa are immensely widely used by *NAS implementations. But they are fundamentally designed for server environments, with forced A/C cooling, not home servers that may have ordinary case+CPU fans.
They are designed to run hot and usually do. Very, very, very hot in home server scenarios.
The 4-8 device cards (1-2 ports+fanout) like the 9211 are usually okay. But it's widely documented that cards with 3+ ports (12 up to maybe 24 devices) routinely overheat causing kernel panic and shutdown, because the cards simply can't cope under passive cooling, even at idle.
A big problem is, the temp isn't exposed in the GUI. These cards routinely anbd happily operate at 90C +, but they cause panic and shutdown at very little more usage, perhaps 107-110C according to my system. When a panic happens, the user is simply told "mpr sensor 0 temp 106C...107C... shutdown/reboot", nothing more.
Given the wide use of these cards in home + small office servers and other non-forced cooling scenarios, combined with their propensity to overheat+panic on idle let alone heavier use if >2 ports, it would be extremely useful to provide a way for the HBA card temps to be monitored and reported/alerted, as a graph, on the dashboard/HBA status, and in logs if needed.
SInce the OS clearly can read these temps correctly (as shown by console etc alerts), they shouldn't be hard to expose in the GUI appropriately. It would be very helpful and give people an idea how close to the wind their system is getting to a panic+shutdown due to card heating, and/or how much margin they have.
Hopefully not a big ask but really important for users with cards over 2 ports.