Jump to content

Recommended Posts

Posted

I am trying to investigate why borg backups, specifically when doing borg check, are taking over twice as long on armbian buster + OMV5, compared to armbian stretch + OMV4. The borg server is an Odroid HC2, and borg check runs on the server (HC2) and is single threaded.

 

Not sure where to even start, but one thing that caught my attention is that the borg process will be running at 100% CPU at various times (as expected), but the CPU frequency at those same times was reported at 200-300 MHz when observed using htop. Is this normal? Unfortunately I don't know what frequency was normal in the same circumstances with armbian stretch + OMV4.

 

Benchmarks done by armbian-config show the CPU at 2000 MHz or a bit less as the passively cooled CPU throttles a bit over extended loads:
 

Throttling statistics (time spent on each cpufreq OPP) for CPUs 4-7:

2000 MHz: 3526.66 sec
1900 MHz:    4.97 sec
1800 MHz:    9.55 sec
1700 MHz:    9.22 sec
1600 MHz:   37.34 sec
1500 MHz:   73.44 sec
1400 MHz:   77.32 sec
1300 MHz:    5.47 sec
1200 MHz:       0 sec
1100 MHz:       0 sec
1000 MHz:       0 sec
 900 MHz:       0 sec
 800 MHz:       0 sec
 700 MHz:       0 sec
 600 MHz:       0 sec
 500 MHz:       0 sec
 400 MHz:       0 sec
 300 MHz:       0 sec
 200 MHz:         sec

I can provide the output of armbianmonitor -U or parts of it if it would help, though at 17000+ lines, pastebin didn't like it. There seem to be a bunch of duplicated sections. armbianmonitor -u said "Network/firewall problem detected. Not able to upload debug info."

Posted

I finally figured this out: there is no problem with CPU frequency, rather, htop is reporting only LITTLE core frequencies for all 8 (4 big, 4 LITTLE) cores. Checking with armbianmonitor -M shows the big cores were at 2000 MHz when a process is at 100% on cpus 4-7, while htop was showing 200-300 MHz.

Guest
This topic is now closed to further replies.
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines