The IBM Serveraid M5014 and M5015 were the flagship product of their time.
They do suffer from overheating and short battery life, depending on how they are installed as their high heat output combined with often inadequate cooling hastens their demise.
Often the heat-sink retaining spring loaded clips will break due to prolonged high temperature, leading first to the raid controller dropping in performance, and eventually totally failing. Renewing the thermal compound and using nylon screws to replace the clips while retaining the springs can resurrect the card if caught in time.
Old cards can be impossible to firmware update without several intermediate versions, and finding them can be hard.
I have one with the original release firmware that MegaCLI doesn't detect. There is an intermediate update required as well, but this one has proved a significant challenge, and still remains stuck at 12.0.1-0065.
I have maintained various models of IBM storage since the DS3500 series, then moving to the Storwize V7000 and V3700
The V3700 has been interesting, as a linux based system, especially addressing issues due to boot ssd bad blocks and file corruption. After more than 10 years working on them, I have become quite skilled in reviving them from a failed state. Having created several clean base disk images of the boot ssd, it is easy to return it to a known candidate state ready as a spare.
I'll post more details here.
Every manufacturer has the odd paperweight in their line and a standout was the Seagate Barracuda 3000GB model ST3000DM01
Of an origional 50 I installed, 90% have failed with smart warnings. There was a firmware upgrade released, but it didn't help reduce the failure rate, so I flagged the model as junk, and replaced them all to prevent potential data loss.