flower

Rockchip says CPU is Ok with temps up to 80 degree. Do you have any load? Without load my unit is around 50 degree. if i stress test cpu and hdd it goes up to 78 degree. think about airflow when installing that small fan. normally the air goes from the front to the backside. another fan would probably disturb that flow. regarding emmc: the speedup is not that big. actually i prefer running of sd because it is way easier to backup and to replace. there are not many writes to the emmc while running anyway (not sure if this is true for omv though) EDIT: you can also install updates without any fear while using sd. just keep a known working copy somewhere

Some of you had the problem that the back of the enclosure does not really fit and stuck a few mm to the outside - leading to loose usb cables. I use noctua fans which are bigger and had the same problem since then. I could solve it by putting that fan cage thingy outside and not inside the case. Maybe it helps you too Gesendet von meinem CLT-L29 mit Tapatalk

ls -alh /dev/fan-p? lrwxrwxrwx 1 root root 41 Oct 15 00:00 /dev/fan-p6 -> /sys/devices/platform/p6-fan/hwmon/hwmon5 lrwxrwxrwx 1 root root 41 Oct 15 00:00 /dev/fan-p7 -> /sys/devices/platform/p7-fan/hwmon/hwmon4 maybe its just the symlink thats missing?

i think its pwm_fan. hard to tell as /dev/fan is a symlink and doesnt have a major:minor number root@ghost:~# lsmod|grep pwm_fan pwm_fan 20480 0

Good news 5.8 seems stable now. I did a paralell stress test on cpu and all hdds for an hour and had it running for about 10h since without any problems. 4.4 would lock. Only missing part for me is auto poweron as soon as my house power is back up and 2.5gbit perf problems. I put my unit in production now. Gesendet von meinem CLT-L29 mit Tapatalk

just donated a little. thank you for this project!

only through a fresh install? i would have preferred f2fs. for flash memory it has a way better durability. anyway... thank you. i wont write much there anyway - all data is in docker volumes on hdd

I downclocked the cpu to 1.6ghz and constrained unrar (which caused my crashes) to two cores. I am on 5.8 I am stable since. Maybe its time for a new cpu and hdd stress test to see if it works now regardless of load. Gesendet von meinem CLT-L29 mit Tapatalk

Not yet. They are working on it

You can do that through /etc/fancontrol but the syntax is a litte strange. I tweaked mine to start the fans earlier (i have replaced the original ones with two 80mm noctua which are more silent) cat /sys/class/hwmon/hwmon0/pwm1 104 cat /sys/class/hwmon/hwmon1/pwm1 89 Shows current fan speed. Values are between 0 and 255 FCTEMPS connect cpu temp to both fans. My /etc/fancontrol INTERVAL=10 FCTEMPS=/dev/fan-p6/pwm1=/dev/thermal-cpu/temp1_input /dev/fan-p7/pwm1=/dev/thermal-cpu/temp1_input MINTEMP=/dev/fan-p6/pwm1=45 /dev/fan-p7/pwm1=45 MAXTEMP=/dev/fan-p6/pwm1=70 /dev/fan-p7/pwm1=70 MINSTART=/dev/fan-p6/pwm1=60 /dev/fan-p7/pwm1=60 MINSTOP=/dev/fan-p6/pwm1=29 /dev/fan-p7/pwm1=29 MINPWM=20

usually it is better to use the batteries charging level instead of a fixed time value as a trigger. eg below 40% shutdown.

if someone wants to shutdown their unit on powerloss: create file: /etc/udev/rules.d/15-gpio-charger.rules SUBSYSTEM=="power_supply", ENV{POWER_SUPPLY_ONLINE}=="0", RUN+="/usr/sbin/halt" i'd love to know a way to start that unit when power is available again though. on 4.4 the system crashes on halt. but as it does this after all services are stopped and the disks are synced this is not really a problem. on 5.8 the system stops and "system halted" appears on serial console - BUT it is still on and fans are spinning.

if its the disks fault you might see it in smartlog smartctl --all /dev/sda

Just a few more questions (i am on kernel 5.8): - is there a way to see if the battery is used? i'd like to write a service which shuts down the unit on powerloss - i get log messages about tx disconnect for r8152. followed by usb errors. iperf shows my transferrate correctly at 2.5gbits but file transfers hover around 80-150mb/s which is a bit slow usb 2-1.4: reset SuperSpeed Gen 1 USB device number 3 using xhci-hcd r8152 2-1.4:1.0 eth1: Tx status -2 xhci-hcd xhci-hcd.0.auto: bad transfer trb length 16744300 in event trb r8152 2-1.4:1.0 eth1: Tx status -2 xhci-hcd xhci-hcd.0.auto: bad transfer trb length 16744300 in event trb r8152 2-1.4:1.0 eth1: Tx status -2 xhci-hcd xhci-hcd.0.auto: bad transfer trb length 16744300 in event trb r8152 2-1.4:1.0 eth1: Tx status -2 xhci-hcd xhci-hcd.0.auto: bad transfer trb length 16744300 in event trb r8152 2-1.4:1.0 eth1: Tx timeout i just want to know the reasoning behind this two: - why does armbian use zram for logs? AFAIK zram is not swappable (instead of tmpfs). - why is zram used for swap instead of zswap? i have always used zswap on my desktop helio64 sets 2gb as swap through zram. sounds like a good value. i have added 4x2gb swap partitions with a lower priority. is this a good idea with the way armbian handles ram (eg swappiness=100)?

Is there a way to update from legacy to current? I am on legacy atm and have nearly finished that setup. As soon as 2.5gbit perf problems are solved i would like to switch though. If not i would reinstall now with current and hope the perf problem will solve itself over time. Btw i can only second ebin-dev. I am really impressed by the quality of this board and case. I really love it.

thank you with this version i am able to build my raid10 array - its already at 5%. the old one locked at 0.3% - raid5 and 6 worked fine though (throuch cmd and omv) i have tried docker with omv which worked fine on the old image (omv installed through softly, docker installed through omv). but i plan to go without omv.

Is it possible to update through apt or would you recommend to flash that new image?

thank you; will do that.

Hello, i've preordered an Helios64 and would like to install the main os on a f2fs formatted sd-card. your documentation says that this is not possible, and i would love to see this possibility. i guess most eMMC devices would also benefit