[priyeshkaratha]

What changes were proposed in this pull request?

Adds a Bucket Level Pending Deletion tab to the Recon OM Insights page, built on the APIs added in
HDDS-17503.

The table lists, per bucket, the total replicated bytes pending deletion (tooltip shows the key and directory
counts and sizes), sorted largest first, with the usual limit selector and volume/bucket search.

Snapshot trapped bytes are calculated on demand, because the OM-side calculation is expensive:

• The Calculate button opens a confirmation dialog explaining the load on OM before submitting.
• While the job is queued/running, the cell shows a status tag and the job is polled every 3s.
• On completion the row shows Snapshot Trapped Bytes, plus columns for Snapshot Trapped Keys,
  Snapshot Trapped Directories, Purgeable Bytes, Purgeable Keys and Purgeable Directories
  (- until the calculation completes). The tooltip on the trapped bytes shows when OM calculated it, and a
  recalculate button re-submits (with the same warning).
• On failure it shows a "Failed" tag with the error and a Retry button. A job that Recon no longer tracks
  (404, e.g. after a Recon restart) is shown as failed.
• On load, recent jobs are fetched so earlier results are shown without recalculating.

Mock API data (db.json, routes.json) is added for local UI development.

Depends on HDDS-17503 for the backend APIs.

What is the link to the Apache JIRA

HDDS-17504

How was this patch tested?