I've got an issue where a simple script using hddtemp is not being polled/captured by the spine poller (I think).
The script takes approximately 2.5 seconds to complete
Code: Select all
root@wdhq:/usr/share/cacti/site/scripts# time ./hdd_temps.sh
hddtemp_sda:55.0 hddtemp_sdb:53.0 hddtemp_sdc:53.0 hddtemp_sdd:52.0 hddtemp_sde:53.0 hddtemp_sdf:54.0
real 0m2.383s
user 0m0.009s
sys 0m0.059s
Code: Select all
01/06/2015 07:15:01 AM - SYSTEM STATS: Time:0.3310 Method:spine Processes:4 Threads:4 Hosts:5 HostsPerProcess:2 DataSources:28 RRDsProcessed:17
As a result, the graph for that rrd shows values of zero (was previously showing NaN).
Any ideas why the poller isn't waiting for the script to return data before finishing?