search

cloud-bulldozer / benchmark-comparison

Project to help compare benchmark result data and system metric data provided from benchmark run.
Other
1 stars 23 forks source link

How the comparison deal with the date in the ES data? #52

Closed inevity closed 3 years ago

inevity commented 3 years ago

Same as Title

rsevilla87 commented 3 years ago

Hi @inevity, not sure what you mean with your question, I'll try to answer it with some comments: Benchmark-comparison is able to output metric aggregations such as avg, max P99 from a given set of documents with the same id (uuid by default). We recently have added the feature of querying and outputing raw data, documented at https://github.com/cloud-bulldozer/benchmark-comparison#querying-for-raw-data, by using that feature you can get a whole series of documents in different formats.

inevity commented 3 years ago

For example, we have two uuid data, such as es time date ingest by the snafu wrapper.
we just metric aggregations such as avg, max P99. I now understand this. Should we need the two uuid data should be same time duration, or should have same amount of the data point to compare?

rsevilla87 commented 3 years ago

For example, we have two uuid data, such as es time date ingest by the snafu wrapper. we just metric aggregations such as avg, max P99. I now understand this. Should we need the two uuid data should be same time duration, or should have same amount of the data point to compare?

Hey @inevity, they don't need to be 100% similar, those documents just need to have the fields in common you use for buckets and/on filters. i.e, find more examples in the config dir. https://github.com/cloud-bulldozer/benchmark-comparison/tree/master/config