Windows add-on 8.0.0, Splunk 8.0.4.
No matter the interval settings in inputs.conf, they seem to run at random times. For example on one host alone, the "service" checker ran 9 times in one hour. The setting in the stanza is for once a day (86400). I've tried setting it other values -- nothing seems to matter.
Also happening on all other inputs (sourcetype=WinHostMon) that have an interval setting.
Disk, for example (also set interval = 86400) is running 2-16 times for host in one hour.
I've searched for this, and heard the descriptions of the scripts "taking a long time to run" yadda, yadda... but come on, not all of them...and these aren't scripts (and we have arguably over powered hardware running this). This is generating a *lot* of entries for our small test group of only 200.
Thoughts?
Thanks.
Mike
It all looks ok. Unless there is something that's doing an overide. can you try using btool to check if inputs are all correct.
Great suggestion on using btool...
However, it confirmed that all is good. I'm focusing on [WinHostMon://Disk] for my test case.
I definitely have:
[WinHostMon://Disk]
interval = 86400
disabled = 0
type = Disk
And definitely confirmed my system is showing data for my Name="C:" at exactly 2 hour intervals... but once in a while only waits an hour in between. Go figure.
What are the inputs.conf settings for the respective inputs and where are they set? If you use universal forwarders then the settings must be on the UFs. Be sure to restart Splunk after changing inputs.conf settings.
[WinHostMon://Disk]
interval = 86400
disabled = 0
type = Disk
...straight out-of-the-box -- (except that I enabled it and changed the interval). In these cases it's being handed out by a deployment server to UFs. Yes, I know the inputs.conf is being applied because other changes are reflected. There's only one app being applied. Yes, it's in local.
Yes, I know to restart/reload it... 😕