[New Feature]: Upgrades to Tosca UI

[sjlewis-jpl]

Checked for duplicates

No - I haven't checked

Alternatives considered

Yes - and alternatives don't suffice

Related problems

I'm frustrated when using the Bach UIs to analyze or assess testing results, and am limited by the interface and display options.

Describe the feature request

This is a list of improvements to the Tosca UI in the browser. These would definitely enhance our ability to monitor and analyze testing results, and I suspect would also greatly improve the experience for regular operations. There is a lot here, and I don't expect it would all be addressed together. The goal here is to document all of these potential improvements together, and the team can pick off when/how to do each.

First suggestion (and I know this might be inherited from HySDS and not changeable): give this UI a meaningful name, one that gives some clue as to why to use the UI, or what the UI does for a user. Happy to brainstorm on ideas if this is actually feasible.

• Timestamp filters: allow all of them to select both the date and the time. Currently one can only select the date. This greatly limits the ability to do testing, as tests run on the same day get mixed together. (adding times to the filters was already submittedin [Bug]: Figaro and Tosca datetime filters don't allow filtering by time #839)
• New temporal filters that would be really useful:
  • Production Time (if this is the same as the "Created At" filter, then rename it to "Production Time")
  • Sensor Time
• New temporal filters that would be "nice to have":
  • PGE job queued time (useful when matching with trigger logic logs)
  • PGE execution time
• Other new filters for granules that would be really useful:
  • Granule identifier: e.g. the burst ID, or MGRS Tile ID, etc. Obviously, these are too numerous to list out, even in a drop-down list. But allowing the user to input them as a list could work.
  • Granule Group identifier: e.g. the MGRS TileSet ID for DSWx-S1
  • Filter by products NOT delivered to the DAAC. Currently we can filter for those already delivered, but we almost always want to see the opposite.
  • Triaged jobs for a specific product type, rather than just all triaged jobs as their own product type.
• Customizable table columns: being able to select, from a list, which columns to show on the table. The list should at least contain any field that can be filtered on with the preset filters on the left. Also, start_time and end_time columns are only populated for triaged jobs, and are usually just wasting space in the UI.
• Issues with / Improvements to existing filters (on the left hand side):
  • "Created At" filter: non-intuitive to use, since none of the columns seem to obviously relate.
  • "Start Time" and "End Time" filters: these only seem to function for triaged jobs, is that right? Specifying that in the UI would be nice.
    • Also, given that these are in Tosca, it's not clear if these times refer to the PGE job that's triaged, or the sensing times for the impacted granule, or something else. Triaged jobs in the data product UI seems odd, and like they'd have a more natural home in Figaro.
    • If we do keep the triaged jobs here, I suggest changing the column names to something like "pge_job_start_time" (which is what I think they're currently capturing?).
    • Have these columns populated for the rest of the data products as well.
  • The "Dataset" and "Dataset Type" filters seem completely redundant. I suggest removing one, or making them distinct.
  • For "Dataset Level": the "NA" option really suggests that Tosca should also be tracking ancillary file inputs. I think that would be really useful going forward, given the dynamic ancillaries in release 3.
    • Also, I think it might make more intuitive sense for the triaged jobs to take the Dataset Level of their outputs (and/or their inputs).
  • System Version: this one is mislabled. It's actually filtering on the Product Version, so I suggest we just change the name in the UI. (already reported in [Bug]: Tosca UI "System Version" filter isn't filtering on the System Version #841)
• Ability to export the tables to CSV format. Exporting all of the results would be very useful for analysis in Excel or Python. Also being able to export just the "current page" (of 100 results, say) could also be useful.
• For "Page Size": options for larger numbers, say up to 1000 (this would just be a "nice to have" if the export to Excel was implemented and fast).
• Button at the top to go directly to the Reporting UI. There's one to jump to Figaro. One to jump to the Reporting UI would be really useful (this is a "nice-to-have").

Many, if not all of these, might be achievable with the right elasticsearch query string, which is already available. However, the syntax to use isn't obvious, and in particular it's difficult to figure out the field name options. If the information button next to this field (little blue circled "i") could redirect to a page specific to our PCM and Elasticsearch database, that could go a long way towards many of these items.