Hello,
I've just upgraded from pi3 to pi5 and it seems a lot faster.
Of course, by adding third party hardware, i have some issues with the peripherals. Here is my problem:
- Using hardware setup from [1]
- Mounted both ssds
- After several minutes, hdds become very hot. Temperature reaches up to 85 degrees.
- Probably kernel tries to throttle or shut them down.
- After a while, hdds are shut down and no longer available (until next reboot).
Despite of this, when hdds stay cool (sometimes more than hour at about 50 degrees C), speeds are good enough (400MB write speed). I've even tried software raid and it was working fine (with mdadm), until the overheating appears.
I've observed that at the time when hdds stop working, the dmesg error appears [2].
Post became very long, but anyone that could propose some solution for the overheating and the hdd shut down is welcome.
One solution is the heatsync, but it won't help much. I'll add it anyways, but I need some software changes in order to stop the overheating.
Thanks
[1]
- 2 x Samsung 990 EVO 2TB (with latest firmware according to Samsung Magician). Only mounted and formatted as ext4/ntfs (tried both).
- X1005 board by GeekWorm
- Raspberry Pi 5, 8GB with latest firmware (2024/10/21)
[2]
[ 3666.916855] nvme nvme0: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0xffff
[ 3666.916864] nvme nvme0: Does your device have a faulty power saving mode enabled?
[ 3666.916866] nvme nvme0: Try "nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off" and report a bug
[ 3666.976855] nvme 0000:03:00.0: Unable to change power state from D3cold to D0, device inaccessible
[ 3666.976874] nvme nvme0: Disabling device after reset failure: -19
[ 3728.356665] nvme nvme1: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0xffff
[ 3728.356673] nvme nvme1: Does your device have a faulty power saving mode enabled?
[ 3728.356676] nvme nvme1: Try "nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off" and report a bug
[ 3728.440678] nvme 0000:04:00.0: Unable to change power state from D3cold to D0, device inaccessible
[ 3728.440700] nvme nvme1: Disabling device after reset failure: -19
I've just upgraded from pi3 to pi5 and it seems a lot faster.
Of course, by adding third party hardware, i have some issues with the peripherals. Here is my problem:
- Using hardware setup from [1]
- Mounted both ssds
- After several minutes, hdds become very hot. Temperature reaches up to 85 degrees.
- Probably kernel tries to throttle or shut them down.
- After a while, hdds are shut down and no longer available (until next reboot).
Despite of this, when hdds stay cool (sometimes more than hour at about 50 degrees C), speeds are good enough (400MB write speed). I've even tried software raid and it was working fine (with mdadm), until the overheating appears.
I've observed that at the time when hdds stop working, the dmesg error appears [2].
Post became very long, but anyone that could propose some solution for the overheating and the hdd shut down is welcome.
One solution is the heatsync, but it won't help much. I'll add it anyways, but I need some software changes in order to stop the overheating.
Thanks
[1]
- 2 x Samsung 990 EVO 2TB (with latest firmware according to Samsung Magician). Only mounted and formatted as ext4/ntfs (tried both).
- X1005 board by GeekWorm
- Raspberry Pi 5, 8GB with latest firmware (2024/10/21)
[2]
[ 3666.916855] nvme nvme0: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0xffff
[ 3666.916864] nvme nvme0: Does your device have a faulty power saving mode enabled?
[ 3666.916866] nvme nvme0: Try "nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off" and report a bug
[ 3666.976855] nvme 0000:03:00.0: Unable to change power state from D3cold to D0, device inaccessible
[ 3666.976874] nvme nvme0: Disabling device after reset failure: -19
[ 3728.356665] nvme nvme1: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0xffff
[ 3728.356673] nvme nvme1: Does your device have a faulty power saving mode enabled?
[ 3728.356676] nvme nvme1: Try "nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off" and report a bug
[ 3728.440678] nvme 0000:04:00.0: Unable to change power state from D3cold to D0, device inaccessible
[ 3728.440700] nvme nvme1: Disabling device after reset failure: -19
Statistics: Posted by jshan — Thu Oct 31, 2024 9:27 am