I am trying to monitor my NVIDIA GPUs on my linux machines. lm-sensors is not available for my motherboard so I setup a script to pull the data.
nvidia-smi --query-gpu=clocks.gr,clocks.sm,clocks.mem,clocks.video --format=csv,nounits,noheader |
awk -v fields='clocks_gr,clocks_sm,clocks_mem,clocks_video' -F', ' '
BEGIN{ split(fields, a, ",") }
{ for (i=1;i<=NF;i++) printf a ":" $i (i==NF ? ORS : " ") }'
output example:
clocks_gr:1995 clocks_sm:1995 clocks_mem:4215 clocks_video:1845
RRDtool Command:
/usr/bin/rrdtool graph - \
--imgformat=PNG \
--start='1587906960' \
--end='1587993360' \
--pango-markup \
--title='GPU Clock Speeds' \
--vertical-label='MHz' \
--slope-mode \
--base=1000 \
--height=150 \
--width=500 \
--alt-autoscale-max \
--lower-limit='0' \
COMMENT:"From 2020/04/26 08\:16\:00 To 2020/04/27 08\:16\:00\c" \
COMMENT:" \n" \
--border 1 --slope-mode \
--watermark 'Generated by Cacti®' \
DEF:a='/usr/share/cacti/site/rra/slowmo_clocks_gr_105.rrd':'clocks_gr':AVERAGE \
LINE2:a#00FF00FF:'' \
GPRINTLAST:'%8.2lf %s'
RRDtool Says:
OK
The problem I am having is that the data Cacti is getting is not what the script is outputting. In this example, the clocks_gr should have a value of about 1995, but the cacti graphs numbers between 2 - 17ish...
NVIDIA GPU stats - problem
Moderators: Developers, Moderators
Who is online
Users browsing this forum: No registered users and 3 guests