Optimize CLUSTER BY for statement_info and metric tables

[aptend]

Optimize CLUSTER BY for statement_info and metric tables

What type of PR is this?

• API-change
• BUG
• Improvement
• Documentation
• Feature
• Test and CI
• Code Refactoring

Which issue(s) this PR fixes:

Related to https://github.com/matrixorigin/MO-Cloud/issues/5327

What this PR does / why we need it:

This PR optimizes the CLUSTER BY configuration for statement_info and metric tables to improve query performance and fix zonemap filtering issues.

Changes

1. statement_info table: Changed CLUSTER BY from using account (varchar) to (request_at, account_id)
2. metric table: Changed CLUSTER BY from (account, metric_name, collecttime) to (collecttime, account_id, metric_name)

Problem

The previous CLUSTER BY configuration used account (varchar(1024)) as the first clustering key. This caused issues because:

• Long account names can fill up the zonemap, making it ineffective for filtering
• String-based clustering is less efficient than integer-based clustering
• Time-based queries are common but time fields were not prioritized in the clustering order

Solution

• Replace account (varchar) with account_id (int) in CLUSTER BY to avoid zonemap overflow
• Place time fields (request_at/collecttime) first in the clustering order to optimize time-range queries
• This ensures zonemap filtering works correctly and improves query performance for time-based and account-based filters

Implementation Details

• Added upgrade scripts in pkg/bootstrap/versions/v4_0_0/ to migrate existing tables
• Updated table definitions in pkg/util/trace/impl/motrace/schema.go and pkg/util/metric/mometric/metric.go
• Updated related query builder logic to handle the new clustering configuration
• Added test cases to verify the changes