Logo

Contents:

  • Getting Started
    • Preparing Your Machine to Install SQreamDB
    • Installing SQreamDB
    • Executing Statements in SQreamDB
    • Performing Basic SQream Operations
      • Running the SQream SQL Client
      • Creating Your First Table
      • Listing Tables
      • Inserting Rows
      • Running Queries
      • Deleting Rows
      • Saving Query Results to a CSV or PSV File
    • Hardware Guide
    • Staging and Development Hardware Guide
  • Installation Guides
    • Installing and Launching SQreamDB
      • Pre-Installation Configuration
      • Installing SQream Using Binary Packages
      • Installing Monit
      • Launching SQream with Monit
      • In-Process Python Functions
    • Installing SQream Studio
      • Installing Studio on a Stand-Alone Server
      • Installing an NGINX Proxy Over a Secure Connection
    • Upgrade Guides
      • Version Upgrade
      • Upgrade-Related Configuration Changes
  • SQreamDB on AWS
  • Operational Guides
    • Access Control
      • Overview
      • Password Policy
      • Managing Roles
      • Permissions
      • Departmental Example
    • Accelerating Filtered Statements
    • Creating or Cloning Storage Clusters
    • Working with External Data
    • Foreign Tables
    • System Health Monitoring
    • Deleting Data
    • Logging
    • Query Split
    • Monitoring Query Performance
    • Security
    • Saved Queries
    • Optimization and Best Practices
    • Oracle Migration Guide
  • Configuration Guides
    • Configuring SQream
      • Cluster and Session
      • Workers
      • Modification Methods
      • Parameter Values
      • Showing All Flags in the Catalog Table
    • LDAP
    • Single Sign-On
  • Architecture
    • Internals and Architecture
    • Filesystem and Usage
    • Sizing
  • Acceleration Studio
    • Getting Started with SQream Acceleration Studio
    • Executing Statements and Running Queries from the Editor
    • Viewing Logs
    • Creating, Assigning, and Managing Roles and Permissions
    • Configuring Your Instance of SQreams
  • Connecting to SQreamDB
    • Client Platforms
      • Denodo Platform
      • Informatica Cloud Services
      • MCP | Integrating SQream DB and Anthropic Claude
      • MicroStrategy
      • Pentaho Data Integration
      • PHP
      • BI Desktop
      • R
      • SAP BusinessObjects
      • SAS Viya
      • Semarchy
      • SQL Workbench
      • Tableau
      • Talend
      • TIBCO Spotfire
    • Client Drivers
      • SQreamNET
      • Dataiku
      • JDBC
      • Node.JS
      • ODBC
        • Install and Configure ODBC on Windows
        • Install and configure ODBC on Linux
      • Python (pysqream)
      • Spark
      • Trino
  • Data Ingestion Sources
    • Overview
    • Avro
    • CSV
    • Parquet
    • ORC
    • JSON
    • SQLoader As a Service
      • Preparing Oracle for Data Migration
  • External Storage Platforms
    • Azure Blob Storage
    • Google Cloud Platform
    • HDFS Environment
    • Amazon Web Services
  • Feature Guides
    • Query Healer
    • Compression
    • Python User-Defined Functions
    • Workload Manager
    • Concurrency and Locks
    • Data Encryption
      • Overview
      • Encryption Methods
      • Data Types
      • Syntax
      • Examples
      • Limitations
      • Permissions
  • References
    • SQL Statements and Syntax
      • SQL Syntax Features
      • SQL Statements
      • SQL Functions
        • Built-In Scalar Functions
        • User-Defined Functions
        • Aggregate Functions
        • Window Functions
    • Catalog Reference
      • Catalog Tables
      • Additional Tables
      • Examples
    • Command line programs
      • metadata_server
      • sqreamd
      • Multi-Platform Sqream SQL
      • Server Picker
      • SqreamStorage
      • sqream sql
      • upgrade_storage
    • SQL Feature Checklist
  • Data Types
    • Supported Data Types
      • Primitive Data Types
      • Array
    • Casts and Conversions
    • Supported Casts
      • Numeric
      • Boolean
      • Integer
      • Floating Point
      • String
      • Date
      • DateTime2
  • Release Notes
    • SQDB release policy
    • 4.x Release Notes
      • Release Notes 4.3
      • Release Notes 4.5
      • Release Notes 4.8
      • Release Notes 4.9
      • Release Notes 4.10
      • Release Notes 4.11
      • Release Notes 4.12
      • Release Notes 4.13
      • Release Notes 4.14
      • Release Notes 4.15
      • Release Notes 4.16
      • Release Notes 4.17
      • Release Notes 4.18
  • Troubleshooting
    • Remedying Slow Queries
    • Resolving Common Issues
    • Identifying Configuration Issues
    • Lock Related Issues
    • Log Related Issues
    • Core Dumping Related Issues
    • Retrieving Execution Plan Output Using SQreamDB Studio
    • Gathering Information for SQream Support
  • Glossary
SQreamDB
  • SHOW STUCK UNLINKS

SHOW STUCK UNLINKS

SHOW_STUCK_UNLINKS lists groups of deleted files that SQreamDB has tried to remove from disk and failed at least once.

When you drop a table, or clean up after a DELETE, the files that held the data are not removed at once. They are queued, and a background process removes them from disk. If removing a group of files fails, for example because of a storage error, the group is queued again and retried. Use SHOW_STUCK_UNLINKS when disk space is not freed after a DROP TABLE or a clean-up, to see whether any files are waiting on a failed removal and where they are.

See also Deleting Data, DELETE.

Syntax

show_stuck_unlinks_statement ::=
    SELECT SHOW_STUCK_UNLINKS()
    ;

The function takes no arguments.

Returns

A result set with one row for each group of files whose removal has failed at least once. Files that are queued for removal but have not yet been tried are not listed. When no removal has failed, the result set is empty.

Result columns

Column name

Description

path

The folder that holds the files

files

The number of files in the group

attempts

The number of times removing the group has failed

key

The metadata key under which the group is queued

Permissions

The role must have SUPERUSER permissions.

Notes

  • A large set of files from one folder is queued in several groups, so the same path can appear in more than one row.

  • The attempts count is stored in the metadata, so it is kept across a restart.

  • The result is capped at 100,000 rows. If more groups are stuck, the rest are not shown and a warning is written to the log.

  • The function reads the queue across all databases in the cluster.

Examples

Listing files whose removal has failed

SELECT SHOW_STUCK_UNLINKS();

© Copyright 2026 SQreamDB.