[BUG] ClickHouse always inserts into a Distributed table, breaking single-node deployments

## Description
`initClient0` always executes `CREATE_DISTRIBUTED_TABLE_SQL` using `config.getClusterName()`, and `consume0` always inserts into `request_log_distributed`. On a single-node ClickHouse (no cluster configured), `clusterName` is null -> `Distributed('null', ...)`, the distributed table creation fails (caught and logged), so `request_log_distributed` is never created, and every batch insert then fails.

## Location
```
shenyu-plugin-logging-clickhouse/.../client/ClickHouseLogCollectClient.java:140, 87
ClickHouseLoggingConstant.java:68 (PRE_INSERT_SQL targets request_log_distributed)
```

## Impact
On a single-node ClickHouse deployment without a `clusterName` configured, all log batches throw on insert; `AbstractLogCollector` catches and drops them, so no logs are persisted.

## Suggested fix
Only create/insert-through the distributed table when `clusterName` is configured; otherwise insert directly into `request_log`.

## Related existing issue(s)
None

_Identified during the 2026-08-02 audit; full list in [`docs/issue-candidates-2026-08-02.md`](docs/issue-candidates-2026-08-02.md)._

Start with ClickHouseLogCollectClient.java at initClient0 and consume0, then inspect ClickHouseLoggingConstant.java:68 where PRE_INSERT_SQL targets request_log_distributed. Trace how config.getClusterName() selects table creation and insertion. Done means single-node deployments insert into request_log without a cluster name, while configured clusters continue using the distributed table.

Written by the indexing model from the issue text.