Dynamic Property Reload | Apache Ozone
Skip to main content
Apache Ozone
Docs
Latest Docs
Docs 2.0.0 (Archived)
Docs 1.4.1 (Archived)DownloadRoadmapBlog
Community
Communication Channels
Who Uses Ozone?
Report An Issue
Ask a Question
How to Contribute
Events and Media
Search
Version:
2.1.1
Next
2.2.0
2.1.1
Overview
Quick Start
Core Concepts
User Guide
Administrator Guide
Installation
Configuration
Operations
Start and Stop
Upgrade and Downgrade
Decommissioning and Maintenance
Disk Replacement
Container Replication Report
Data Balancing
Certificate Rotation
S3 Multi-Tenancy
Snapshots
Observability
Leader Transfer
Quota in Ozone
Tools
Trash
Dynamic Property Reload
System Internals
Developer Guide
This is documentation for Apache Ozone 2.1.1, which is no longer actively maintained.
For up-to-date documentation, see the latest version (2.2.0).
Administrator Guide
Operations
Dynamic Property ReloadVersion: 2.1.1
On this page
Dynamic Property Reload
Ozone supports dynamic reloading of certain configuration properties without restarting services. This enables operators to tune cluster behavior, adjust limits, and update settings in production without service disruption.
Overview​
When a property is marked as reconfigurable, you can:
Modify the property value in the configuration file (ozone-site.xml)
Invoke the reconfig command to apply the changes to the running service
The reconfiguration is performed asynchronously, and you can check the status to verify completion.
Command Reference​
ozone admin reconfig --service=[OM|SCM|DATANODE] --address=<ip:port|hostname:port> <operation>
Options​
OptionDescription--serviceThe service type: OM, SCM, or DATANODE--addressRPC address of the target server (e.g., hadoop1:9862 or 192.168.1.10:9862). Required unless --in-service-datanodes is specified.--in-service-datanodes(Datanode only) Apply to all IN_SERVICE Datanodes
Operations​
OperationDescriptionstartExecute reconfiguration asynchronouslystatusCheck the status of a reconfiguration taskpropertiesList all reconfigurable properties for the service
Reconfigurable Properties​
Ozone Manager (OM)​
PropertyDefaultDescriptionozone.administrators-Comma-separated list of Ozone administratorsozone.readonly.administrators-Comma-separated list of read-only administratorsozone.om.server.list.max.size1000Maximum server-side response size for list operationsozone.om.volume.listall.allowedtrueAllow all users to list all volumesozone.om.follower.read.local.lease.enabledfalseEnable local lease for follower read optimizationozone.om.follower.read.local.lease.lag.limit10000Maximum log lag for follower readsozone.om.follower.read.local.lease.time.ms5000Lease time in milliseconds for follower readsozone.key.deleting.limit.per.task50000Maximum keys to delete per taskozone.directory.deleting.service.interval60sDirectory deletion service run intervalozone.thread.number.dir.deletion10Number of threads for directory deletionozone.snapshot.filtering.service.interval60sSnapshot SST filtering service run interval
Storage Container Manager (SCM)​
PropertyDefaultDescriptionozone.administrators-Comma-separated list of Ozone administratorsozone.readonly.administrators-Comma-separated list of read-only administratorshdds.scm.block.deletion.per-interval.max500000Maximum blocks SCM processes per deletion intervalhdds.scm.replication.thread.interval300sInterval for the replication monitor threadhdds.scm.replication.under.replicated.interval30sFrequency to check the under-replicated queuehdds.scm.replication.over.replicated.interval30sFrequency to check the over-replicated queuehdds.scm.replication.event.timeout12mTimeout for replication/deletion commandshdds.scm.replication.event.timeout.datanode.offset6mOffset subtracted from event timeout for Datanode deadlinehdds.scm.replication.maintenance.replica.minimum2Minimum replicas required for node maintenancehdds.scm.replication.maintenance.remaining.redundancy1Remaining redundancy required for maintenance (EC)hdds.scm.replication.datanode.replication.limit20Max replication commands queued per Datanodehdds.scm.replication.datanode.reconstruction.weight3Weight multiplier for reconstruction commandshdds.scm.replication.datanode.delete.container.limit40Max delete container commands queued per Datanodehdds.scm.replication.inflight.limit.factor0.75Factor to scale cluster-wide replication limithdds.scm.replication.container.sample.limit100Number of containers sampled per state for debuggingozone.scm.ec.pipeline.minimum5Minimum EC pipelines to keep openozone.scm.ec.pipeline.per.volume.factor1Factor for calculating EC pipelines based on volumes
Datanode​
PropertyDefaultDescriptionhdds.datanode.block.deleting.limit.per.interval20000Maximum blocks deleted per interval on a Datanodehdds.datanode.block.delete.threads.max5Maximum threads for block deletionozone.block.deleting.service.workers10Number of block deletion service workersozone.block.deleting.service.interval60sBlock deletion service run intervalozone.block.deleting.service.timeout300sBlock deletion service timeouthdds.datanode.replication.streams.limit10Maximum replication streams per Datanode
Usage Examples​
List Reconfigurable Properties​
To view all properties that can be dynamically reconfigured:
$ ozone admin reconfig --service=OM --address=hadoop1:9862 properties
OM: Node [hadoop1:9862] Reconfigurable properties:
ozone.administrators
ozone.om.server.list.max.size
ozone.om.volume.listall.allowed
OM Reconfiguration​
Modify ozone.administrators in ozone-site.xml, then execute:
$ ozone admin reconfig --service=OM --address=hadoop1:9862 start
OM: Started reconfiguration task on node [hadoop1:9862].
$ ozone admin reconfig --service=OM --address=hadoop1:9862 status
OM: Reconfiguring status for node [hadoop1:9862]: started at Wed Dec 28 19:04:44 CST 2022 and finished at Wed Dec 28 19:04:44 CST 2022.
SUCCESS: Changed property ozone.administrators
From: "hadoop"
To: "hadoop,bigdata"
SCM Reconfiguration​
Modify ozone.administrators in ozone-site.xml, then execute:
$ ozone admin reconfig --service=SCM --address=hadoop1:9860 start
SCM: Started reconfiguration task on node [hadoop1:9860].
$ ozone admin reconfig --service=SCM --address=hadoop1:9860 status
SCM: Reconfiguring status for node [hadoop1:9860]: started at Wed Dec 28 19:04:44 CST 2022 and finished at Wed Dec 28 19:04:44 CST 2022.
SUCCESS: Changed property ozone.administrators
From: "hadoop"
To: "hadoop,bigdata"
Datanode Reconfiguration​
Modify hdds.datanode.block.deleting.limit.per.interval in ozone-site.xml, then execute:
$ ozone admin reconfig --service=DATANODE --address=hadoop1:19864 start
Datanode: Started reconfiguration task on node [hadoop1:19864].
$ ozone admin reconfig --service=DATANODE --address=hadoop1:19864 status
Datanode: Reconfiguring status for node [hadoop1:19864]: started at Wed Dec 28 19:04:44 CST 2022 and finished at Wed Dec 28 19:04:44 CST 2022.
SUCCESS: Changed property hdds.datanode.block.deleting.limit.per.interval
From: "20000"
To: "30000"
Batch Operations (Datanode Only)​
To perform reconfiguration on all IN_SERVICE Datanodes simultaneously:
$ ozone admin reconfig --service=DATANODE --in-service-datanodes start
Datanode: Started reconfiguration task on node [hadoop1:19864].
Datanode: Started reconfiguration task on node [hadoop2:19864].
Datanode: Started reconfiguration task on node [hadoop3:19864].
Reconfig successfully 3 nodes, failure 0 nodes.
To list properties across all Datanodes:
$ ozone admin reconfig --service=DATANODE --in-service-datanodes properties
Datanode: Node [hadoop1:19864] Reconfigurable properties:
hdds.datanode.block.deleting.limit.per.interval
Datanode: Node [hadoop2:19864] Reconfigurable properties:
hdds.datanode.block.deleting.limit.per.interval
Datanode: Node [hadoop3:19864] Reconfigurable properties:
hdds.datanode.block.deleting.limit.per.interval
Reconfig successfully 3 nodes, failure 0 nodes.
Best Practices​
Test in non-production first: Always validate configuration changes in a test environment before applying to production.
Change one property at a time: When making multiple changes, apply them incrementally to isolate the impact of each change.
Monitor after changes: Watch cluster metrics and logs after reconfiguration to ensure the changes have the desired effect.
Document changes: Keep a record of configuration changes for troubleshooting and audit purposes.
Use batch operations carefully: When using --in-service-datanodes, ensure all nodes should receive the same configuration.
Edit this page
Previous
Trash
Next
System Internals
Overview
Command Reference
Options
Operations
Reconfigurable Properties
Ozone Manager (OM)
Storage Container Manager (SCM)
Datanode
Usage Examples
List Reconfigurable Properties
OM Reconfiguration
SCM Reconfiguration
Datanode Reconfiguration
Batch Operations (Datanode Only)
Best Practices
Apache Software Foundation
Foundation
License
Events
Sponsorship
Privacy
Security
Thanks
Community
GitHub Discussions
Jira Issues
Slack
Mailing List
YouTube
Twitter
Repositories
Ozone
Website
Docker Image
Docker Runner Image
Copyright © 2026 The Apache Software Foundation. Licensed under the Apache License, Version 2.0.
The Apache Software Foundation, Apache Ozone, Ozone, Apache, the Apache Feather, and the Apache Ozone project logo are either registered trademarks or trademarks of the Apache Software Foundation.