Self-Repair should not crash

### Is your feature request related to a problem?

During the phase of self-repair, after listing the transactions to download from the beacon's chain, it may happen that some transactions cannot be fetched for some reason, causing the process to halt, and the node is not responding. This creates a fault-tolerance issue.

### Describe the solution you'd like

In the process of data retrieval between the beacon's chain & the transaction's chain, the node should accept to not be able to recover the transaction's from the chain, present in the beacon. This transaction would be repaired by the live repair of the transaction's chain validation phase.
Then, if the transaction is not able to recovered, the self-repair should skip this case, and continue the flow to ingest more transactions. This would make the self-repair more fault-tolerant.

### Additional context

_No response_

### Epic

_No response_

Start by tracing the self-repair data-retrieval flow between the beacon's chain and the transaction's chain. Verify how an unavailable transaction is handled, then confirm that self-repair skips it, continues ingesting later transactions, and leaves recovery to live transaction-chain validation.

Written by the indexing model from the issue text.