serhiy-bzhezytskyy opened a new issue, #158:
URL: https://github.com/apache/solr-orbit/issues/158

   ### Description
   `convert-workload` writes every `date_histogram` as a range facet with 
`start: NOW/YEAR-10YEAR` and `end: NOW/YEAR+1YEAR`, whatever the data holds. 
The facet moves with the clock, and for a 2015 corpus it matches nothing.
   
   The same conversion also drops two things without failing: `time_zone` 
(every bucket edge moves) and a nested `stats` aggregation (the log says 
`Unsupported aggregation type 'stats' — skipping`).
   
   ### To reproduce
   ```
   solr-orbit convert-workload --workload-path 
<opensearch-benchmark-workloads>/nyc_taxis --output-path out/nyc_taxis
   ```
   In `out/nyc_taxis/operations/default.json`, 
`date_histogram_calendar_interval` has `"start": "NOW/YEAR-10YEAR"`. Posted to 
a Solr 10.1.0 collection holding 300,640 nyc_taxis documents it reports 
`numFound: 300640` and no non-empty bucket; 
`date_histogram_fixed_interval_with_metrics` does the same. 
`date_histogram_calendar_interval_with_tz` and 
`date_histogram_fixed_interval_with_tz` carry no zone, and 
`date_histogram_fixed_interval_with_metrics` carries none of its twelve metrics.
   
   ### Expected behavior
   `start` and `end` come from the query's range on the aggregated field; a 
`fixed_interval` starts where OpenSearch aligns it (60d over 2015 starts at 
2014-11-06); `time_zone` is kept; `stats` becomes min, max, sum and avg. With 
no range in the query the operation should say so, rather than return nothing.
   
   ### Host / Environment
   solr-orbit main e37cc020, Solr 10.1.0.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to