Add calculation-period to reporting elements #279
Conversation
We discussed about this change on Green-Software-Foundation#123 . Add workload-start-time and calculation-period to SCI reporting elements as optional fields. It helps us to understand the impact of start times and processing times of workloads on SCI.
| @@ -22,6 +22,8 @@ The following list contains the REQUIRED and OPTIONAL data elements to be report | |||
| | Software or Product Version | metadata/entity-version | MUST | Text | | | |||
| | SCI Specification Version | metadata/sci-version | MUST | Numeric | The version of the SCI specification against which this report is being made. | | |||
| | Date of Calculation | metadata/calculation-date | MUST | Date | Following a format described in [RFC3339], [a subset](https://ijmacd.github.io/rfc3339-iso8601/) of ISO 8601 | | |||
| | Start date of the workload | metadata/workload-start-time | MAY | Timestamp | Following a format described in [RFC3339], [a subset](https://ijmacd.github.io/rfc3339-iso8601/) of ISO 8601 | | |||
| | Processing period for calculation | metadata/calculation-period | MAY | Numeric | Time taken in seconds to complete the process defined by `R`. | | |||
There was a problem hiding this comment.
Should the time be in ms since some tasks / processes might be much more granular than taking seconds?
There was a problem hiding this comment.
I've updated to use ms instead of sec.
I thought it may be suitable to use sec here because the task period could be long (e.g. ML job). However we should also consider short-term task (e.g. transaction). So I accepted the suggestion from @atg-abhishek .
|
WG: we need to think more carefully about the appendix overall before we address specific issues/questions. |
|
@Henry-WattTime |
|
See issue #283 for resolution of this PR. |
We discussed about this change on #123 . This PR is based on this comment.
Add
workload-start-timeandcalculation-periodto SCI reporting elements as optional fields.It helps us to understand the impact of start times and processing times of workloads on SCI.
We can analyze, and feedback to run the job more efficiency with this PR. So I hope merge this PR to v1.0.