Fix default workspace name for PostgreSQL AGE graph storage
This commit is contained in:
yangdx
parent
bab2803953
commit
80f7e37168
3 changed files with 64 additions and 45 deletions
|
|
@ -882,7 +882,7 @@ rag = LightRAG(
|
|||
|
||||
* **对于Neo4j图数据库，通过label来实现数据的逻辑隔离**：Neo4JStorage
|
||||
|
||||
为了保持对遗留数据的兼容，在未配置工作空间时PostgreSQL的默认工作空间为`default`，Neo4j的默认工作空间为`base`。对于所有的外部存储，系统都提供了专用的工作空间环境变量，用于覆盖公共的 `WORKSPACE`环境变量配置。这些适用于指定存储类型的工作空间环境变量为：`REDIS_WORKSPACE`, `MILVUS_WORKSPACE`, `QDRANT_WORKSPACE`, `MONGODB_WORKSPACE`, `POSTGRES_WORKSPACE`, `NEO4J_WORKSPACE`。
|
||||
为了保持对遗留数据的兼容，在未配置工作空间时PostgreSQL非图存储的工作空间为`default`，PostgreSQL AGE图存储的工作空间为空，Neo4j图存储的默认工作空间为`base`。对于所有的外部存储，系统都提供了专用的工作空间环境变量，用于覆盖公共的 `WORKSPACE`环境变量配置。这些适用于指定存储类型的工作空间环境变量为：`REDIS_WORKSPACE`, `MILVUS_WORKSPACE`, `QDRANT_WORKSPACE`, `MONGODB_WORKSPACE`, `POSTGRES_WORKSPACE`, `NEO4J_WORKSPACE`。
|
||||
|
||||
## 编辑实体和关系
|
||||
|
||||
|
|
|
|||
|
|
@ -928,7 +928,7 @@ The `workspace` parameter ensures data isolation between different LightRAG inst
|
|||
- **For relational databases, data isolation is achieved by adding a `workspace` field to the tables for logical data separation:** `PGKVStorage`, `PGVectorStorage`, `PGDocStatusStorage`.
|
||||
- **For the Neo4j graph database, logical data isolation is achieved through labels:** `Neo4JStorage`
|
||||
|
||||
To maintain compatibility with legacy data, the default workspace for PostgreSQL is `default` and for Neo4j is `base` when no workspace is configured. For all external storages, the system provides dedicated workspace environment variables to override the common `WORKSPACE` environment variable configuration. These storage-specific workspace environment variables are: `REDIS_WORKSPACE`, `MILVUS_WORKSPACE`, `QDRANT_WORKSPACE`, `MONGODB_WORKSPACE`, `POSTGRES_WORKSPACE`, `NEO4J_WORKSPACE`.
|
||||
To maintain compatibility with legacy data, the default workspace for PostgreSQL non-graph storage is `default` and, for PostgreSQL AGE graph storage is null, for Neo4j graph storage is `base` when no workspace is configured. For all external storages, the system provides dedicated workspace environment variables to override the common `WORKSPACE` environment variable configuration. These storage-specific workspace environment variables are: `REDIS_WORKSPACE`, `MILVUS_WORKSPACE`, `QDRANT_WORKSPACE`, `MONGODB_WORKSPACE`, `POSTGRES_WORKSPACE`, `NEO4J_WORKSPACE`.
|
||||
|
||||
## Edit Entities and Relations
|
||||
|
||||
|
|
|
|||
|
|
@ -117,32 +117,44 @@ class PostgreSQLDB:
|
|||
"""
|
||||
|
||||
existing_columns = await self.query(check_columns_sql, multirows=True)
|
||||
existing_column_names = {col['column_name'] for col in existing_columns} if existing_columns else set()
|
||||
existing_column_names = (
|
||||
{col["column_name"] for col in existing_columns}
|
||||
if existing_columns
|
||||
else set()
|
||||
)
|
||||
|
||||
# Add missing chunk_id column
|
||||
if 'chunk_id' not in existing_column_names:
|
||||
if "chunk_id" not in existing_column_names:
|
||||
logger.info("Adding chunk_id column to LIGHTRAG_LLM_CACHE table")
|
||||
add_chunk_id_sql = """
|
||||
ALTER TABLE LIGHTRAG_LLM_CACHE
|
||||
ADD COLUMN chunk_id VARCHAR(255) NULL
|
||||
"""
|
||||
await self.execute(add_chunk_id_sql)
|
||||
logger.info("Successfully added chunk_id column to LIGHTRAG_LLM_CACHE table")
|
||||
logger.info(
|
||||
"Successfully added chunk_id column to LIGHTRAG_LLM_CACHE table"
|
||||
)
|
||||
else:
|
||||
logger.info("chunk_id column already exists in LIGHTRAG_LLM_CACHE table")
|
||||
logger.info(
|
||||
"chunk_id column already exists in LIGHTRAG_LLM_CACHE table"
|
||||
)
|
||||
|
||||
# Add missing cache_type column
|
||||
if 'cache_type' not in existing_column_names:
|
||||
if "cache_type" not in existing_column_names:
|
||||
logger.info("Adding cache_type column to LIGHTRAG_LLM_CACHE table")
|
||||
add_cache_type_sql = """
|
||||
ALTER TABLE LIGHTRAG_LLM_CACHE
|
||||
ADD COLUMN cache_type VARCHAR(32) NULL
|
||||
"""
|
||||
await self.execute(add_cache_type_sql)
|
||||
logger.info("Successfully added cache_type column to LIGHTRAG_LLM_CACHE table")
|
||||
logger.info(
|
||||
"Successfully added cache_type column to LIGHTRAG_LLM_CACHE table"
|
||||
)
|
||||
|
||||
# Migrate existing data using optimized regex pattern
|
||||
logger.info("Migrating existing LLM cache data to populate cache_type field (optimized)")
|
||||
logger.info(
|
||||
"Migrating existing LLM cache data to populate cache_type field (optimized)"
|
||||
)
|
||||
optimized_update_sql = """
|
||||
UPDATE LIGHTRAG_LLM_CACHE
|
||||
SET cache_type = CASE
|
||||
|
|
@ -154,7 +166,9 @@ class PostgreSQLDB:
|
|||
await self.execute(optimized_update_sql)
|
||||
logger.info("Successfully migrated existing LLM cache data")
|
||||
else:
|
||||
logger.info("cache_type column already exists in LIGHTRAG_LLM_CACHE table")
|
||||
logger.info(
|
||||
"cache_type column already exists in LIGHTRAG_LLM_CACHE table"
|
||||
)
|
||||
|
||||
except Exception as e:
|
||||
logger.warning(f"Failed to add columns to LIGHTRAG_LLM_CACHE: {e}")
|
||||
|
|
@ -288,7 +302,6 @@ class PostgreSQLDB:
|
|||
logger.warning(f"Failed to check LLM cache migration status: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def _migrate_llm_cache_to_flattened_keys(self):
|
||||
"""Optimized version: directly execute single UPDATE migration to migrate old format cache keys to flattened format"""
|
||||
try:
|
||||
|
|
@ -330,7 +343,9 @@ class PostgreSQLDB:
|
|||
|
||||
conflict_result = await self.query(conflict_check_sql)
|
||||
if conflict_result and conflict_result["conflicts"] > 0:
|
||||
logger.warning(f"Found {conflict_result['conflicts']} potential ID conflicts with existing records")
|
||||
logger.warning(
|
||||
f"Found {conflict_result['conflicts']} potential ID conflicts with existing records"
|
||||
)
|
||||
# Can choose to continue or abort, here we choose to continue and log warning
|
||||
|
||||
# Execute single UPDATE migration
|
||||
|
|
@ -358,9 +373,13 @@ class PostgreSQLDB:
|
|||
remaining = verify_result["remaining_old"] if verify_result else -1
|
||||
|
||||
if remaining == 0:
|
||||
logger.info(f"✅ Successfully migrated {old_count} LLM cache records to flattened format")
|
||||
logger.info(
|
||||
f"✅ Successfully migrated {old_count} LLM cache records to flattened format"
|
||||
)
|
||||
else:
|
||||
logger.warning(f"⚠️ Migration completed but {remaining} old format records remain")
|
||||
logger.warning(
|
||||
f"⚠️ Migration completed but {remaining} old format records remain"
|
||||
)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Optimized LLM cache migration failed: {e}")
|
||||
|
|
@ -1767,8 +1786,8 @@ class PGGraphStorage(BaseGraphStorage):
|
|||
"""
|
||||
Generate graph name based on workspace and namespace for data isolation.
|
||||
Rules:
|
||||
- If workspace is empty: graph_name = namespace
|
||||
- If workspace has value: graph_name = workspace_namespace
|
||||
- If workspace is empty or "default": graph_name = namespace
|
||||
- If workspace has other value: graph_name = workspace_namespace
|
||||
|
||||
Args:
|
||||
None
|
||||
|
|
@ -1777,15 +1796,15 @@ class PGGraphStorage(BaseGraphStorage):
|
|||
str: The graph name for the current workspace
|
||||
"""
|
||||
workspace = getattr(self, "workspace", None)
|
||||
namespace = self.namespace or os.environ.get("AGE_GRAPH_NAME", "lightrag")
|
||||
namespace = self.namespace
|
||||
|
||||
if workspace and workspace.strip():
|
||||
if workspace and workspace.strip() and workspace.strip().lower() != "default":
|
||||
# Ensure names comply with PostgreSQL identifier specifications
|
||||
safe_workspace = re.sub(r"[^a-zA-Z0-9_]", "_", workspace.strip())
|
||||
safe_namespace = re.sub(r"[^a-zA-Z0-9_]", "_", namespace)
|
||||
return f"{safe_workspace}_{safe_namespace}"
|
||||
else:
|
||||
# When workspace is empty, use namespace directly
|
||||
# When workspace is empty or "default", use namespace directly
|
||||
return re.sub(r"[^a-zA-Z0-9_]", "_", namespace)
|
||||
|
||||
@staticmethod
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue