manage_hypothetical_indexes
Test PostgreSQL query performance by creating hypothetical indexes in memory to evaluate potential improvements without building physical indexes.
Instructions
Manage HypoPG hypothetical indexes for testing.
HypoPG allows you to create "hypothetical" indexes that exist only in memory and can be used to test query plans without the overhead of creating real indexes.
Actions:
create: Create a new hypothetical index
list: List all current hypothetical indexes
drop: Drop a specific hypothetical index
reset: Drop all hypothetical indexes
estimate_size: Estimate the size of a hypothetical index
This is useful for:
Testing if an index would improve a query
Comparing different index strategies
Estimating index storage requirements
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| action | Yes | Action to perform | |
| table | No | Table name (required for create, estimate_size) | |
| columns | No | Column names for the index (required for create, estimate_size) | |
| index_type | No | Type of index to create | btree |
| unique | No | Whether the index should be unique | |
| index_id | No | Index OID (required for drop) |
Input Schema (JSON Schema)
{
"properties": {
"action": {
"description": "Action to perform",
"enum": [
"create",
"list",
"drop",
"reset",
"estimate_size",
"check"
],
"type": "string"
},
"columns": {
"description": "Column names for the index (required for create, estimate_size)",
"items": {
"type": "string"
},
"type": "array"
},
"index_id": {
"description": "Index OID (required for drop)",
"type": "integer"
},
"index_type": {
"default": "btree",
"description": "Type of index to create",
"enum": [
"btree",
"hash",
"gin",
"gist",
"brin"
],
"type": "string"
},
"table": {
"description": "Table name (required for create, estimate_size)",
"type": "string"
},
"unique": {
"default": false,
"description": "Whether the index should be unique",
"type": "boolean"
}
},
"required": [
"action"
],
"type": "object"
}