How to configure the graph text as input?

[hannesd]

Hi all,

I want to create a graph, which I can fill up with the following data sources:

tbs_name: This is the name of the Table-Space
tbs_usage: This is the usage from the Table-Space

My script returns me the following output:

Code: Select all

tbs_name:TBS1 tbs_usage:1 tbs_name:TBS2 tbs_usage:7683 tbs_name:TBS3 tbs_usage:4 tbs_name:TBS5 tbs_usage:14180

I've already created two data-sources: tbs_name & tbs_usage.

How can I now configure the graph, that I can see:
- The tablespace-name (certain color) with his tablespace-usage without having an own graph for every tablespace?

What "Data Sources Type" must the field "tbs_name" have?

My idea is, that I have one graph, where can I see all table-spaces with their usages.

Thanks a lot.
Hannes

[Linegod]

http://docs.cacti.net/manual:087:3a_adv ... ut_methods

[hannesd]

I already saw this document. But I don't know, how to configure the graph, to have all values (the name of the tablespace AND the tablespace-usage) in one graph.

What "Data Sources Type" must the field "tbs_name" have?

Thanks a lot.

[Linegod]

You can't pull back a text value as a Data Source Type.

You would either have to manually input the description as per http://docs.cacti.net/manual:087:3a_adv ... ut_methods or you would have to create an Indexed script as per http://docs.cacti.net/manual:087:3a_adv ... alkthrough and utilize query_ for tbs_name.

However, you would have to differentiate between one variable and another in order to use it within a legend (tbs_name_1, tbs_name_2....)

[Linegod]

Come to think of it, the differentiation would not be necessary if each pair was in a separate data source - similar to host_disk.xml

[hannesd]

mmh....I think this is too complicated for me. In this case, I will make an own data-source for each table-space. I although prefer one rrd with all tablespaces...but - how it seems - it's too complex.

Or can somebody help me with this challenge?

Thanks a lot.