Skip to content

[v9] fix: Use float mapping for fractional service/agent monitoring fields - #8798

Draft
chrisburr wants to merge 1 commit into
DIRACGrid:rel-v9r0from
chrisburr:fix/monitoring-float-mapping-v9r0
Draft

chrisburr wants to merge 1 commit into
DIRACGrid:rel-v9r0from
chrisburr:fix/monitoring-float-mapping-v9r0

Conversation

@chrisburr

Copy link
Copy Markdown
Member

ResponseTime, CpuPercentage and CycleDuration are floats (seconds and percent) but were mapped as long, so OpenSearch truncated them in the indexed values. Aggregations (avg, percentiles) therefore returned 0 for all sub-second response times, while _source kept the correct values.

The new mapping only applies to newly created (weekly) indices; existing indices need a reindex to recover correct aggregations.

BEGINRELEASENOTES

*MonitoringSystem
FIX: Use float mapping for ResponseTime, CpuPercentage and CycleDuration in ServiceMonitoring and AgentMonitoring so that aggregations are no longer truncated (existing indices need to be reindexed)

ENDRELEASENOTES

ResponseTime, CpuPercentage and CycleDuration are floats (seconds and
percent) but were mapped as long, so OpenSearch truncated them in the
indexed values. Aggregations (avg, percentiles) therefore returned 0 for
all sub-second response times, while _source kept the correct values.

The new mapping only applies to newly created (weekly) indices; existing
indices need a reindex to recover correct aggregations.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant