+- Pandora FMS community forums (http://pandorafms.org/forum)
+-- Forum: Pandora FMS (http://pandorafms.org/forum/forum-3.html)
+--- Forum: Advanced troubleshooting and problems (http://pandorafms.org/forum/forum-11.html)
+--- Thread: issues with generic_data_inc (/thread-4486.html)
issues with generic_data_inc - eXXon - 12-19-2011
Hi there!
I have some troubles with data type generic_data_inc:
as says documentation:
Quote:The whole numeric data equals to the differential being between the current value and the previous one. When this differential is negative, the value is fixed to 0.
Im create a modules to monitor eth1 data flow, here the definition of modules:
Code:
# Networking.
module_begin
module_name nettraf_eth1_in
module_type generic_data_inc
module_exec cat /sys/class/net/eth1/statistics/rx_bytes
module_postprocess 0.0009765625
module_description Bytes IN (postprocess ti Kb)
module_end
module_begin
module_name nettraf_eth1_out
module_type generic_data_inc
module_exec cat /sys/class/net/eth1/statistics/tx_bytes
module_postprocess 0.0009765625
module_description Bytes OUT (postprocess to Kb)
module_endbut i get randomly "0" value in data( see screenshot ),
as i know /sys/class/net/eth1/statistics/tx_bytes returns a value of total transmitted bytes on eth1 and this value always increment. But how there zerros appear? and why timestamp of zerro values are same. I look it in mysql
Code:
SELECT * FROM `tagente_datos`WHERE `id_agente_modulo` =%AgentID%the zerro values have same timestamp with the previous one (see screenshot)
Re: issues with generic_data_inc - eXXon - 12-23-2011
Find out problem:
file Core.pm
1860 str
# Should not happen
return 0 if ($utimestamp == $data_inc->{'utimestamp'});
that cases incorrect average data calculations in graphs.
Also to devs - its better in inc data type check if last value = now value don't write in db. Less records in db => better performance.
attached corected file with catch this type of exception;
Re: issues with generic_data_inc - rnovoa - 12-26-2011
Hi eXXon,
Thank you for taking the time to send us a patch. I've fixed the part where 0 was returned when a second value with the same timestamp arrived, but data compression is already handled in the function pandora_process_module, it should work for incremental modules, ¿doesn't it work for you? Thank you again.