The purpose of this RFC is to solicit input on a draft Change Request which will clarify requirements and specifications in the OSS (LSE-30) and the DMSR (LSE-61) that are related to the alert stream.
The proposed changes are provided in the uploaded ascii file.
- is child task of
DM-19483 RFC on exactly how to define OTT1
- relates to
DM-15654 Define policy for handling fields with > 10K alerts
- To Do
DM-20308 Migrate contents of LDM-502 to Jira LVV project
- mentioned in
In OSS-REQ-0193, the specification refers to an 'average value of at least nAlertVisitAvg' whereas the Description field defines nAlertVisitAvg as a 'Minimum number of alerts required to be accommodated from a single standard visit'. This is not consistent, the Description should be updated to specify that nAlertVisitAvg is an average value computed nightly.
In Item 6: the description sates ' The long-term average number of alerts per visit.', but we are computing this average nightly - I would update this to say 'The nightly average number of alerts per visit'
To satisfy Issue E (the mild confusiong between documents about nAlertVisitAvg), we proposed to more tightly define nAlertVisitAvg with a new specification in the DMSR (Item 6).
However, the corresponding (and slightly confusing) specification OSS-REQ-0193 should also be updated so that OSS and DMSR match in their definition of nAlertVisitAvg.
Therefore, Item 7 now proposes updates to OSS-REQ-0193 that clarify the definition of nAlertVisitAvg (in addition to clarifying the term 'degrade gracefully', as previously discussed).
Minor updates to the text of Item 6 were also made to clarify that nAlertVisitAvg is the minimum of a nightly average of the number of alerts per visit, and that this applies to standard visits.
Watchers should please verify that these (minor) updates for items 6 and 7 are satisfactory, and we can move this RFC into a proper LCR.
This RFC is complete, and the Change Request has been submitted:
After discussion in DM-SST telecon today, added "Issue G" and "Item 7" to clarify the phrase "degrade gracefully" in OSS-REQ-0193. This is still under discussion, more feedback still welcome.
Also linked nAlertVisitPeak with OTT1 by altering the opening statement of DMS-REQ-0004 to say "For visits resulting in fewer than nAlertVisitPeak, LSST must be capable of ... " (and added a discussion to sciVisitAlertDelay to make that definition consistent).