Cardano Updates

@vrom911 please take a glance to make sure you're happy with this; but i
think it should fix the static build problems on master.

- Acceptors/Utils.hs: include node_name alongside node_id in SeriesIdentifier
- Component.hs: generalise insert to accept SeriesIdentifier directly; fix
  pruning period default (was 3600 ms ≈ 3.6 s, now 15 min)
- Component/Trace.hs: update TimeseriesTraceInsert to carry SeriesIdentifier;
  emit query_text/query_result only at DDetailed; lower Prune severity to Debug
- TimeseriesServer.hs: rename node/state → node/sync-progress; remove #if RTVIEW
  guard (endpoint is now unconditionally available); return JSON objects from
  GET config/retention and GET config/pruning instead of plain text
- Utils.hs: move NodeStateWrapper (and its FromJSON instance) here from
  RTView/Update/NodeState.hs so it can be shared without the RTVIEW flag;
  move vector from rtview-only to unconditional build-depends
- RTView/Update/NodeState.hs: remove NodeStateWrapper definition (now in Utils)
- grafana-datasource: rename node-state → node-sync-progress throughout
  (QueryEditor, datasource, types, dashboard); rename dashboard to rtview-port
- grafana-datasource: add README with build and run instructions
- grafana-datasource: expose SeriesIdentifier labels as proper Grafana field
  labels instead of baking them into the frame name string; add legendFormat
  field to queries so users can control the legend via {{label_name}} syntax

Signed-off-by: jeluard <[email protected]>

Deletes all artifacts that existed solely to support the external
index-applier container: the Docker Compose file, the apply-indexes.sh
shell script, the Helm Job template, the ConfigMap template and its
bundled db-indexes.yaml, and the two dead _helpers.tpl helpers
(indexApplierName / dbIndexesName).

Also removes jq/yq from the Postgres Dockerfile (only needed for
apply-indexes.sh YAML parsing) and drops the indexApplier resource
blocks from all three Helm hardware profiles and the top-level
indexApplier values section.

Documentation updated across boot-sequence, docker, kubernetes
deployment/helm-values/overview, and index-management pages to
reflect the native yaci-indexer approach.

Co-Authored-By: Claude Sonnet 4.6 <[email protected]>

update code

update code

Co-Authored-By: Claude Sonnet 4.6 <[email protected]>

  - Replace fail-fast `return` with `failedCount` counter so all 16 indexes
    are attempted even when one fails; overall state is READY only if
    failedCount == 0, otherwise FAILED (Tier 1.1)
  - Resolve current schema at init() and qualify all DROP INDEX statements
    as `<schema>.<index>` to avoid cross-schema matches (Tier 1.3)
  - Add guardrail comment: executeIndexCreation always re-queries pg_index
    per-index so any future retry path inherits drop-before-rebuild (Tier 1.2)
  - Update tests: mock current_schema(), fix DROP expectations, add
    continuesAfterSingleFailure test verifying both CREATE calls fire
    despite first failure

Co-Authored-By: Claude Sonnet 4.6 <[email protected]>

feat: Spike: replace 'apply indexes' with a native solution

update code