Management tooling

[lukashornych]

To make evitaLab even more useful, we want to add several tools for database server/catalog/collection management.

For management of database server itself we would like to:

• status 🟢

  • URL, version, API URLs, ... -> solved in Management tooling - Server status #103
  • evitaDB runtime configuration -> solved in Management tooling - Server status #103
  • basic charts? (heap size, cpu usage, ...) -> solved in Management tooling - Server status - Charts #213
  • external links?

• logs + filtering 🟢 -> solved in Management tooling - System logs #142

• JFR tracing (start, stop, download, configuration) 🟢 -> solved in Management tooling - JFR tracing #143

  • optionally embed JFR visualizer if possible

• create catalog -> solved in Management tooling - Creating catalog #144

Cluster management (in future):

• health, lags
• master/replica view

For catalog management we would like to:

• catalog preview -> solved in Management tooling - catalog preview #161
• query recording -> solved in Management tooling - Traffic viewer #150
• time machine 🟢 -> solved in Management tooling - Time machine #95
  • catalog version
    • mutation listing with filtering + mutation detail
    • create new catalog from version
• backup, restore 🟢 -> solved in Management tooling - Backup/restore #147
• drop 🟢
• unload?, load?
• replace, rename 🟢
• (un)make readonly
• create collection
• consistency
• API compatibility
  • REST
  • GraphQL

Entity collection management:

• index listing 🟢
  • size
  • consistency
  • recreate
• drop 🟢
• rename 🟢

Items marked with 🟢 have API support in evitaDB and can be implemented. Each section/item will be solved in separate issue.

[lukashornych]

Maybe we should introduce connection name into toolbars like so

Lots of these tools work on level of evitaDB server/connection and it could be confusing having connection name only in tab name.

@novoj toughs?

[novoj]

Yes, but there is a lot of data to read, and it would be nice to recognize the right connection from the tab name without clicking on it. What about picking for each connection random color and prepend this bullet before catalog name? And do that only when there are more than single connection? Or use another emoji: 1️⃣2️⃣3️⃣ ... etc? It scales up to 10 connections which seems reasonable enough.

[lukashornych]

After some discussion, we will not render connection name in the toolbars (yet). Instead, we will explore different approach to connections entirely in #149.

[lukashornych]

@o-miko Let's focus on following menu items next:

• on connection:
  • create catalog
• on catalog:
  • drop catalog
  • rename catalog
  • replace catalog
  • create collection
• on collection:
  • rename collection
  • drop collection

Let's create these methods in the EvitaDBDriver and implement them using the gRPC API. We could then make ConnectionService.getDriver publically accessible so that UI components can access the correct driver and call required methods. In the future, I think we should use this approach even for existing methods on ConnectionService like getCatalogs, i.e. access is through the driver, because the driver is already the nice API for the evitaLab. The connection service could just provide management for connections and the driver.

[lukashornych]

@o-miko we will add new action to catalog menu > Switch to transactional mode which will appear only if the catalog has state === CatalogState.WARMING_UP. It will call EvitaSessionService.goLiveAndClose gRPC method.
To test this, you will need to create new catalog for it.

[lukashornych]

I've tested the dialogs and we need to:

• fix UI (button placement, sizing and colors, window padding, text wrapping) to match other dialogs (maybe create common dialogs to simplify the process)
• handle exceptions by displaying toaster error notifications
• when adding new collection, UI is not updated
• forms in dialogs miss input validations
• if dangerous dialog is closed via ESC, confirmation dialogs is showed anyway

Otherwise it's working well.

[lukashornych]

Things we need to sort out that we couldn't make in time:

• job type, must change the behavior of the job listing
• can't import catalog from server
• we don't import catalog from local file
• all/none in event selection for record jfr
• add metadata to job files - title, size, etc.