grep

Same problem with "externally powered" HDD (I use 2-plug USB cable and power connected to separate 2.1A USB-charger; cannot post armbianmonitor -u "500 Internal server error"): So, @Igor @guidol @zador.blood.stained +1 to "this is not underpower caused by HDD connection"

@zador.blood.stained I suppose simultaneous run a lot of IO/CPU-intensive tasks (monerod & bitcoind threads) is like a stress-test. But, there is no stability (latest armbianmonitor -u = http://sprunge.us/hUZM ). And line "reset high-speed USB device number 2 using ehci-platform" becomes a bit annoying. What causes "USB device reset" ? Is underpower the only reason ? Or "most likely reason because we haven't seen things like this before ?". I ordered "reference/original" PSU (it did not arrived yet) but i still think the problem is not with PSU (i used 2 different and got the same error). Looks like next step is going to be "using external power for usb hdd" as @guidol adviced, and, if error will happen once again, it will reject all args for assumption "cpu underpower because of usb device".

And once again (now i used BTRFS, armbianmonitor -u = http://sprunge.us/agje ): PSU is this one: https://ru.aliexpress.com/item/Mini-switching-power-supply-15W-5V-3A-Single-Output-dc-power-supply-for-Led-Strip-AC110V/32800882848.html The new one did not arrived yet. But "something" makes me think that problem is not related to PSU.

@guidol @chwe Thanks. Ordered one at AliExpress, but need to wait for delivery.

Same symptoms after reboot (icmp-availability with unability to boot from HDD). In fact this is strange: there was no such behavior before last crash. And it was ok, after i connected it to BOOTED system. Probably: something wrong had happended with board. But looks like first of all i need to replace PSU to overcome "not enough power on boot" possible problem. I have one more (not usual "black cube with plug", but this one https://ru.aliexpress.com/item/Mini-switching-power-supply-15W-5V-3A-Single-Output-dc-power-supply-for-Led-Strip-AC110V/32800882848.html) but its TOO hot (really ~70-80 celcius deg. under load). Can anyone help to find reliable & proven & not too hot 5V/3A PSU ?

@guidol Looks like you're right: power led was not "bright red", but "muted red", i've replaced USB 3.0 cable (i use external case with 3.0 interface) with USB 2.0 dual-plug cable hoping that "theoretically" (it's just an assumption) it could use power from both USB-lines attached. And it is working now. e2fsck -f did not gave any errors. So, do you think "2-plug USB cable" will help me, or there is a "common power circuit" for all USB on this board and i need external power supply to use for USB-HDD ?

@Igor http://sprunge.us/NJFS Looks like my USB-HDD fs is corrupted, but when i connected device with this fs to "single" usb port on board i could not see the device at all Is it normal ?

@Igor i've made a reboot and after that board became unaccessible via ssh (while there's still icmp-connectivity ) it's weird a bit. Need to check and fix.

Just for information (i know there is no support): under heavy load i have such error (for 3 times on different re-runs)

@Igor i put a 19x19x24 small-tower-like radiator for CPU (not too effective, but the best ones i got without "manual working with saw & rasp") and 9x9x12 for each mem + ethernet chips. So if you give me link to testing guide i could try. Of course, i suppose you have "well-defined QA-procedures" and "sporadic results from unknown guys" is not a too good idea, but ... maybe, i just want to burn my board As i "understood" (i guess) from "brief search" the problem is with setting correct DVFS levels (voltage + frequency/throttling) to keep system stable. Simple FREQ=1008Mhz testing with gave [max...min] = [45..49.1] (don't ask about median/avg/quantiles, it was just watching at "armbianmonitor -m" output); normal temp (non-idle, 1.25..2 LA) is ~35..37 P.S. PSU is 5V / 3A of course.

Hi! Could you please clarify, is there a way to increase default limitations for max cpu freq (H5) on Orange Pi Prime? As i understand, according to driver (cpufreq-dt): /etc/default/cpufrequtils is a bit less restrictive: Is this a driver limitation ? Or there is a way to increase cpu-freq, but it's not used by default ? ( ... and why, if it is so ?)