[eXXon]

Hi there!

I have some troubles with data type generic_data_inc:
as says documentation:

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_end

but 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)

[eXXon]

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;

[rnovoa]

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.