Currently, to customize HAProxy backend health check intervals and failover behavior, users must provide a complete HAProxy configuration via haproxy.configuration, duplicating all the operator's default logic and making maintenance difficult.

Problem

When a PXC node fails (e.g., during rolling restart), HAProxy takes 20+ seconds to detect the failure with default settings:

• check inter 10000 (10s interval)
• fall 2 (2 failures = 20s total)

Worse, existing client connections to the failed backend are NOT terminated, causing them to hang until TCP timeout (potentially minutes), as on-marked-down shutdown-sessions is not configured.

Current Workaround

Provide full HAProxy configuration in haproxy.configuration field, but this:

• Duplicates operator logic
• Breaks on operator upgrades
• Is difficult to maintain

Requested Feature

Add CR fields to configure HAProxy backend health check parameters without full config override:

haproxy:
  healthCheck:
    interval: 3000          # check inter 3000 (default: 10000)
    fall: 2                 # fall 2 (default: 2)
    rise: 1                 # rise 1 (default: 1)
    shutdownOnMarkDown: true # on-marked-down shutdown-sessions (default: false)

Benefit

• Fast failover (6s with interval: 3000)
• Active connection termination on backend failure
• No maintenance burden
• Survives operator upgrades

Alternative

If granular options are too complex, at least expose a "fast-failover" preset that applies sensible defaults.