Performance data (may aggregate) in XI

[axvers]

Hi Support,

Besides basic monitoring services (CPU/Memory...), we also set a service monitoring query result.

The problem is: The original "integer" results are shown in "float" on Nagios XI performance data graph.

For example, for the first 5 min, let's say T1, the query result is: 1.0;
the next 5 min, T2: 3.0;
T3: 5.0

On the performance graph, it showed some float value like 0.6 or 1.63 during T1~T3.

My questions are:
A-1. Is that a feature of rrd value?
A-2. Or, is Nagios XI always doing some kind of aggregation when showing performance data?

B-1. Is there a way to disable this? Just show in integer, maybe? (Want T1 is just 1, T2 is just 3...; and zero is zero... etc.)


Thanks!

[scottwilkerson]

This is a feature of how RRDtool works. It averages and compresses the data over time at different step intervals the further away you get from when the data was collected.

RRDtool assumes time-variable data in intervals of a certain length. This interval, usually named step, is specified upon creation of an RRD file and cannot be changed afterwards. Because data may not always be available at just the right time, RRDtool will automatically interpolate any submitted data to fit its internal time-steps.

The value for a specific step, that has been interpolated, is named a primary data point (PDP). Multiple PDPs may be consolidated according to a consolidation function (CF) to form a consolidated data point (CDP). Typical consolidation functions are average, minimum, maximum.

After the data has been consolidated, the resulting CDP is stored in a round-robin archive (RRA). A round-robin archive stores a fixed number of CDPs and specifies how many PDPs should be consolidated into one CDP and which CF to use. The total time covered by an RRA can be calculated as follows:

Code: Select all

time covered = (#CDPs stored) x (#PDPs per CDP) x (step time length)

After this time the archive will "wrap around": the next insertion will overwrite the oldest entry. This behavior in this context is referred to as "round-robin" and is the reason for the program's name. However this is different from the common computer science definition, which is a method of distributing resources among multiple consumers or processes.

To cover several timespans and/or use several consolidation functions, an RRD file may contain multiple RRAs. The data retrieval function of RRDtool automatically selects the archive with the highest resolution that still covers the requested timespan. This mechanism is also used by RRDtool's graphing subsystem.

[axvers]

So, there is no setting for not aggregating the performance data, right?

[lmiltchev]

No, because RRDTool always averages and compresses the data over time at different step intervals. You can review the settings in the /usr/local/nagios/etc/pnp/rra.cfg. Here's our "Performance Data Averaging" KB article:

https://support.nagios.com/kb/article/n ... g-768.html

Importatnt: I would highly advise against modifying the "default" settings in the rra.cfg file, as you may run out of disk space, and end up with a broken system.

[bheden]

Although untested with XI, this should be possible by changing the consolidation functions CF configured in /usr/local/nagios/etc/pnp/rra.cfg.

You'd need to adjust AVERAGE to whichever function you would like to use. I believe you're looking for LAST - and then you would need to delete all existing performance graphs to recreate the round robin databases.

Again, this is untested but should be possible. If you do it, please report back your findings - I'm interested in the result.

[lmiltchev]

Thanks @bheden!

@axvers, if you decide to go this route, make sure you try it in a test environment first, before applying it to production. This hasn't been tested by us.