Skip to content

Trend time series catalog - maybe don't need to cache or can reduce catalog size? #11

@smalers

Description

@smalers

The TSTool Zabbix plugin reads global data when it starts, including a catalog of history time series and a separate catalog of trend time series. The latter assumes that every history time series has Avg, Min, and Max trend time series. However, this is probably not the case because trends are only computed on numerical values (I think). To streamline memory use, can probably do one or more of the following:

  1. Don't use memory for trend catalog but generate it on the fly. I have constrained the queries to require at least one of the data type or a Where filter so generating on the fly may not be too bad.
  2. If the item value type is checked for history catalog, the number of trend time series may be reduced. I'm not sure how much without trying. If the above approach (1) is used, then dynamically created trend catalog records would need to check the item value type.

I'm not sure how much variation there is in value type in order to generate the unique list of data types in TSTool. Need to try one or both of the above and see how it goes.

Metadata

Metadata

Assignees

Labels

SSize: day or lessenhancementNew feature or requesthighPriority: next release if possible

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions