Data source returns correct value, graph still Nan

[burek123]

- SNMP from console returns correct raw value
- Data source is added in GUI
- Data source is added to graph items
- Cacti log shows correct output value

Graph still Nan

SNMP raw value from console:

root@pihole: snmpget -c public -v 2c localhost hrSystem.8
HOST-RESOURCES-MIB::hrSystem.8 = Gauge32: 54230 (this value is divided by 1000 via CDEF function)

Data source template:


Graph template:


Cacti log file, output is correct:
2018/05/13 18:30:03 - POLLER: Poller[Main Poller] Device[11] Device[Pihole] Graphs[RPI CPU temperature, RPI CPU temperature] DS[Pihole - RPI CPU temperature] SNMP: v1: 192.168.0.6, dsname: rpi_temp, oid: hrSystem.8, output: 54768

Cacti graph log file:
2018/05/13 18:30:03 - POLLER: Poller[Main Poller] CACTI2RRD: /usr/bin/rrdtool update /var/www/html/rra/pihole_temp_196.rrd --template rpi_temp 1526229003:U



Any ideas how to solve this?

[netniV]

Are there any values in the graph debug? If all that shows are NaN, this would suggest that the poller isn't polling for the graph.

[burek123]

Graph debug:

/usr/bin/rrdtool graph - \
--imgformat=PNG \
--start='-180000' \
--end='-300' \
--pango-markup \
--title='RPI CPU temperature' \
--vertical-label='Celsius' \
--slope-mode \
--base=1000 \
--height=150 \
--width=500 \
--alt-autoscale-max \
--lower-limit='0' \
--color BACK#F3F3F3 \
--color CANVAS#FDFDFD \
--color SHADEA#CBCBCB \
--color SHADEB#999999 \
--color FONT#000000 \
--color AXIS#2C4D43 \
--color ARROW#2C4D43 \
--color FRAME#2C4D43 \
--border 1 --font TITLE:11:'Arial' \
--font AXIS:8:'Arial' \
--font LEGEND:8:'Courier' \
--font UNIT:8:'Arial' \
--font WATERMARK:6:'Arial' \
--slope-mode \
--watermark 'Generated by Cacti®' \
DEF:a='/var/www/html/rra/pihole_temp_196.rrd':'rpi_temp':AVERAGE \
CDEF:cdefa='a,1000,/' \
LINE1:cdefa#C04000FF:'Temperature' \
GPRINT:cdefa:LAST:'Current\:%8.2lf %s' \
GPRINT:cdefa:MIN:'Min\:%8.2lf %s' \
GPRINT:cdefa:MAX:'Max\:%8.2lf %s\n'
RRDtool Says:
OK

Data source info:

[burek123]

I have removed some other default generic SNMP queries for that host and it suddenly started to work. I still don't know what was the cause ...

[camerabob]

Did you check the poller logs when the query was failing? Perhaps the timeout value was too small and Cacti stopped looking for a response before it was returned? Just throwing stuff at the wall to see what sticks.

[brononi]

Hello,

I'm getting the same issue here with a new installation.
What I've done?

- Installation of Proxmox Virtual machine (=Ubuntu 18.04 LTS)
- Installation of cacti with 'apt install cacti' (= cacti 1.1.38)
- Add the local machine (= with his own IP 192.168.111.4)
Snmp v2, gets nicely connected.
- Add it with the 'local linux Machine' Template
- Add all graphs for this one (Linux Memory Usuage, Unix Load Average, Unix Logged in Users, Unix Processes, Unix Mounted Partitions).

And only the Load Average is being filled. All other graphs stay empty. :$

I've done this installation 2 times from scratch, just to check I've something went wrong during the installation on my side...
In the first installation, I've changed some things in order to get rid of the error you can see below. But that didn't solved the 'nan-value' issue.

In the second installation, I didn't change anything on OS level or cacti. So everything is 'out of the box'...


A snmpwalk:

snmpwalk.txt

(332.18 KiB) Downloaded 95 times

An example of the views:

graphs.png (401.54 KiB) Viewed 2914 times

An example of the debug:

Debug_Memory.png (110.44 KiB) Viewed 2914 times

A screenshot of the clog:

clog.png (459.14 KiB) Viewed 2914 times

I've also added several remote servers, and with the same result. The standard graphs like snmp interface aren't being filled in. With another cacti server in the same network, I don't have these issues... The other one is running since several months, but is on the same level of OS and cacti version. In time, I've done this installation manually.

[cigamit]

The local linux machine templates don't use SNMP. You will want to use different graph templates to utilize SNMP vs the local scripts that they are running. You can also increase the logging level to see what the local scripts are doing.

[brononi]

I just added an external network interface, and it seems to getting the values now...
Strange, yesterday, this resulted in all nan-values.

I'll try to add some additional once...

[netniV]

Sorry, I'm not clear on whether there is still a problem or yesterday there was but today there isn't?

[ozgurerdogan]

I am having same issue... Graphs are emtpy...