op-node: add follow source reorg observability

[jelias2]

Description

This pull request adds new metrics and logging to improve observability around reorg events in the op-node component. It introduces Prometheus counters to track "follow source" and "super authority" reorg events, and integrates these metrics into key reorg-handling code paths. Additional logging is also added to provide more context during reorgs.

Metrics instrumentation and logging improvements:

• Added two new Prometheus counters in metrics.go: FollowSourceReorgs (tracks follow source reorg decisions by action) and SuperAuthorityReorgSignals (tracks super authority reorg signals by reason). Corresponding methods were added to the Metricer interface and implemented in both the real and noop metricers. (op-node/metrics/metrics.go
• In engine_controller.go, integrated RecordSuperAuthorityReorgSignal calls at points where the super authority safe head is unknown or non-canonical, to record the reason for these reorg signals. (op-node/rollup/engine/engine_controller.go)
• In the FollowSource method, added detailed warning and info logs to capture the state of local and external heads during reorgs, and instrumented the code to record follow source reorg actions (force_reset for sequencer resets and soft_update for verifier updates) using the new metric.
  These changes make it easier to monitor and diagnose reorg-related events in the system by providing structured metrics and richer logs.

Tests

Additional context

Metadata