June 2025
PowerScale OneFS [Link] Release Notes
These release notes contain descriptions of the new features, resolved issues, and known issues that are in this version of
OneFS.
Current Release Version: [Link]
Release Type: Major (MA)
Topics:
• Revision history
• Product description
• New features
• Resolved issues
• Known Issues
• Where to get help
Revision history
This section provides a description of document changes.
Table 1. Revision history
Document revision Date Description
1.0 April 8, 2025 Initial release version.
1.1 April 14, 2025 Adding support for PA110 nodes.
1.2 May 19, 2025 Adding support for 61TB and 122TB drives.
1.3 June 23, 2025 Adding support for A310, A3100, H710, and H7100 nodes.
Product description
The scale-out NAS storage platform combines modular hardware with unified software to harness unstructured data. The
OneFS operating system powers your clusters to deliver a scalable pool of storage with a global namespace.
For information about supported hardware and software, see the PowerScale OneFS Supportability and Compatibility Guide on
the Dell Technologies Support site.
New features
This section summarizes the new features in OneFS [Link].
Drive support for new PowerScale A310, A3100, H710, and H7100 nodes support hard drives (HDDs) with capacities of 2 TB,
PowerScale A310, 4 TB, 8 TB, 12 TB, 16 TB, 20 TB, and 24 TB, as well as cache drives (SSDs) with capacities of 800 GB,
A3100, H710, and 1.6 TB, 3.2 TB, and 7.68 TB.
H7100 nodes.
1
Support for new OneFS [Link] and later supports the new PowerScale A310, A3100, H710, and H7100 hardware.
PowerScale A310,
A3100, H710, and
H7100 nodes.
Support for new OneFS [Link] and later supports the new PowerScale PA110 hardware.
PowerScale PA110
node
SmartSync File-to- The SmartSync File-to-Object Backup to Cloud feature expands the one-time File-to-Object Copy to
Object Backup Cloud capacity by introducing a new full-fidelity File-to-Object Backup to Cloud format. This backup
feature added format preserves metadata including encodings, attributes, alternate data streams, hardlinks, and
softlinks through backups, incremental backups, restores, and daisy chain workflows.
ESRS has been ESRS can no longer be configured or enabled, and the end-of-life Secure Remote Services (SRS) v3.x
deprecated gateways are no longer supported by OneFS.
If ESRS is active upon an upgrade commit, and if ESRS is using a v5.x or higher secure connect
gateway, all relevant ESRS configuration options will be migrated to Dell Technologies connectivity
services, which will become active in its place.
The ESRS configuration options will remain functional while the cluster is being upgraded.
Performance Improved performance in large cluster, mixed workload environments, and improved performance
enhancements around S3.
S3A has been added S3A is a Hadoop compatible file system that allows big data applications to store and access data in an
as a supported object storage system as if it were a traditional Hadoop Distributed File System (HDFS). PowerScale
protocol would now support S3A clients. S3A clients would not work with OneFS when the etag is not MD5.
PowerScale recommends using direct S3A uploads, because S3 provides an atomicity guarantee.
SmartConnect SmartConnect Advanced is included with OneFS as a standard feature and is autoenabled. The
Advanced has SmartConnect Basic bundle has been removed.
been included with
PowerScale OneFS
Rolling upgrades have Parallel and simultaneous upgrades are both supported. Parallel upgrades are the recommended
been deprecated upgrade path.
APEX File Storage for APEX File Storage for Azure brings PowerScale OneFS into the public cloud and provides users the
Azure supports IPv4- same management experience as an on-premises PowerScale cluster. IPv4-only or mixed mode IPv4
only or mixed mode and IPv6 modes are now supported. IPv6-only is not supported.
IPv4 and IPv6
Resolved issues
This section describes issues that are resolved in this release of PowerScale OneFS.
API
PSCALE-244190 This fixes the timeout value of papi getting a response from the diagnostic server to 5 seconds.
CloudPools
PSCALE-229671 This fix resolves an issue where incorrectly recalled files continued to be retried for processing without
being marked as Failed.
2
Data Mobility
PSCALE-260106 This fix resolves a core of the isi_migr_sworker process when handling cloud-based objects.
PSCALE-254921 This fix resolves an issue where SMB R2R symbolic links synchronized with SyncIQ were not available
on a target cluster.
PSCALE-251952 This fix resolves an issue where CloudPools archived data or retrieved data could have become
unreadable after the Logical Node Number (LNN) was changed for a node.
PSCALE-247652 This fix resolves an issue where a large file checkpoint created during an alternate data stream
replication during an initial sync may have caused SyncIQ to not transfer some data for a file.
PSCALE-247218 If a SmartSync to Object job pauses with the "too many open files" error, jobs may not pause and an
EMFILE error may be logged. The following configuration value should be incrementally increased until
the job can complete without pausing: isi_dm config set fs::max-open-files.
The following configuration value can also be incrementally decreased: isi_dm config set
task-runner::max-sparks. Adjusting these values decreases the amount of concurrency used
for jobs and decreases the number of open files.
PSCALE-239870 SmartSync job failures may occur if there are active SmartSync jobs when performing an upgrade
from OneFS versions 9.4 through 9.7. This occurs when running jobs are interrupted by multiple
SmartSync daemon service restarts during the upgrade. Resume paused jobs with the following
command: isi dm jobs modify <job_id> --action RESUME_JOB.
Events, Alerts, and Cluster Monitoring
PSCALE-248163 This fix resolves an issue where SNMPv3 traps could stop being accepted by an SNMP receiver after
a cluster or node reboot.
PSCALE-247203 This fix resolves an issue where email alerts failed to send from a cluster and caused a "list index out
of range" error in isi_celog_reporting.log.
PSCALE-244093 This fix resolves an issue where the custom alerting email template was not applied.
File System
PSCALE-254338 This fixes an issue where creating a SmartQoS workload under a specific username would fail if the
zone name was not also specified.
PSCALE-251787 This fix resolves an issue where some cluster capacity could be leaked due to some reference locks on
files and directories not being released properly.
PSCALE-247004 This fix resolves an issue where InsightIQ or API handlers could not retrieve FSA data if a job engine
gconfig setting ([Link].du_split_count) was increased past its default of 10. Users would see an
error: too many attached databases - max 10.
PSCALE-246655 This fix addresses an issue where running the isi get command in the /ifs partition without
additional parameters would return an error.
PSCALE-238064 This fix resolves an issue where jobs would create shared memory but would not clean the memory
when the job finished, creating memory leaks. This fix also adds a more explicit command line for the
isi_usysctl and a cron script to perform garbage collection every hour.
PSCALE-235644 This fix resolves an issue where a QAC imbalance resulted in locking and impacted write performance.
3
PSCALE-229740 This fix resolves an issue that caused an event spam for SW_PP_STATS_QUERY_FAILED event
types.
PSCALE-222209 This fix resolves an issue where isi_post_ifs might have incorrectly created an empty folder,
resulting in lock contention on multiple nodes.
PSCALE-63051 This fix resolves an issue where creating a process core would also update atime, which could
interrupt cluster access when that gcore data is needed for further issue diagnosis.
Hardware
PSCALE-258822 This fix resolves an issue where the interface between OneFS and the iDRAC could stop responding.
PSCALE-254863 This fix validates new settings to prevent errors generated by IPMI Remote Management network
configuration commands.
PSCALE-250434 This fix resolves an issue where a chassis fan failure event failed to report again when the issue
persisted.
PSCALE-244308 This fix resolves an issue where a F600 node may have panicked on SAS controller resets.
PSCALE-242003 This fix adds a --noipmi option to isi_reformat_node which preserves IPMI settings on a node
during a node reformat.
Networking
PSCALE-255673 This fix resolves an issue where the Flexnet process might not have sent gratuitous ARPs during some
IP address moves between interfaces within the same node.
PSCALE-250256 This fix resolves an issue where a node could hang or panic due to a bug within the TCP SYN caching
system.
PSCALE-247613 This fix resolves a cause of excessive critical event spam for backend fabric errors.
PSCALE-244001 This fix prevents CELOG critical event generation for DNOS SmartFabric validation errors whenever
the management interface for a switch is administratively disabled.
PSCALE-241731 In some circumstances, DNS requests to SmartConnect might time out. This change creates separate
threads for SmartConnect DNS processing to avoid any unnecessary latency in responses.
Protocols
PSCALE-258431 This fix resolves an issue where ListObjects generated too many entries in the directory cache, which
led to system instability.
PSCALE-258381 This fix resolves an issue where the copy-object would not copy a directory into another directory.
PSCALE-256908 Fixed issue that AbortMultipartUpload fails if the parent directory of the target file is removed.
PSCALE-253636 This fix resolves an issue where SW_AUDIT_PERSIST_FAILED events could overwhelm the CELOG
analysis daemon and generate considerable amounts of log spam and [Link] bloat.
PSCALE-253390 This fix resolves an issue where the NfsStatsOperationTracked function was not implemented properly
for NLM.
PSCALE-250426 Patched SQL vulnerability where an authenticated attack on user-defined policies could cause
information disclosure.
4
PSCALE-248640 This fix resolves an issue where custom NFSv3 maximum read and write sizes would not be reflected
on the client.
PSCALE-244178 This fix resolves an issue where high NFSv3 read workflow would slow down the NFSv3 write
workflow.
PSCALE-243822 This fix resolves an issue that could cause multiple node panics and an out-of-memory condition in a
cluster.
PSCALE-239170 This fix resolves an issue where a customer may have been unable to restore an NFS configuration
backup containing some unresolvable hosts or export paths.
PSCALE-239149 This fix resolves an issue with the count of [Link] becomes anomalous when the
cluster time is modified, which is due to monotone time used to query data from the kernel and
isi_stats_d was not aware of the time change.
PSCALE-229090 Using less secure authentication methods like gss-ntlm instead of kerberos for RPC endpoint lookup
can cause intermittent disruptions rather than a smooth failover. An SMB client connecting to a CA
share may be unable to register for witness notifications, delaying a quick and seamless failover.
However, this issue resolves itself once the client retries and reconnects.
PSCALE-225485 This fix resolves an issue where a buildup of suspended NfsPersistentRegister fibers caused a node to
reach OOM condition during failover.
PSCALE-218821 This fix resolves an issue where the NFS service may not have started properly due to the zone
refresh thread being blocked by NfsKvstoreActuatorWork.
Remote Connectivity
PSCALE-254914 This fix resolves an issue where a node's /var partition could fill due to isi_rice_d log entries going to
an unlinked isi_rice_journal_pruner.log file.
PSCALE-246496 The 'ese' user that is included in the default OneFS 'System' file is a system-level user. User-
configured auth provider sources must avoid naming a user 'ese'. If a name conflict for the 'ese'
user occurs, Dell Technologies connectivity services may encounter problems with initial provisioning
and with sending telemetry and events to Dell.
Workaround: Rename the 'ese' username to another username within any external auth provider to a
username that does not conflict with any usernames included in the 'System' file provider, as well as
any other OneFS default auth provider. Ensure the renamed username also does not conflict with the
list of 'Unlistable Users' as defined in each OneFS default auth provider.
PSCALE-235447 This fix improves an error message when performing an ESE hard reset.
Upgrades
PSCALE-250744 This fix resolves an issue where rpcbind failed to start correctly during an upgrade, which may have
impacted services needed to register with rpcbind in order to function.
OneFS WebUI
IME01B-17 This fix resolves an issue where the modification of the Quota notification rule value in the WebUI sets
the other rule values to the default value.
PSCALE-246722 This fix resolves an issue where cloud retention settings were not set properly in the OneFS WebUI.
5
Known Issues
This section describes issues that are known to affect this release of OneFS.
For information about issues that are known to affect other OneFS releases, log in to the PowerScale OneFS Product Page and
view the available OneFS Release Notes.
Data Mobility
PSCALE-260403 There is a known issue with creating SmartSync dataset creation policies on a basepath that is a
symbolic link. If such a dataset creation policy is configured, you will receive the following error while
listing the datasets.: "Internal error occurred while accessing the dataset list."
Workaround: Do not attempt to create a dataset creation policy on a basepath that is a symbolic link.
Create the dataset policy on the absolute base path.
PSCALE-224108 There is a known limitation that file paths with mixed encodings cannot be copied to cloud providers. If
a file path with mixed encodings is encountered, the copy job will fail.
Events, Alerts, and Cluster Monitoring
PSCALE-263133 Help text on some of the isi event channels commands indicate that new "connectemc" or
"connectivity" channels can be created or existing channels modified. This text is incorrect, the
system creates these channels and any attempt to manually create or modify them fails.
Networking
PSCALE-262566 When two separate interfaces that support RDMA are connected to two different switches that are
not connected to each other, RDMA clients are not able to connect to one interface and pings from
the node to the clients fail.
Workaround: Either LAGG the two switches together or create a unique subnet for each switch.
Protocol
PSCALE-262679 During an unfortunate contention towards an SMB oplock/lease and then the waiting mechanism
waits for the lock, but hits an error like ENOTCONN (connection to our lock coordinator got
disconnected), then it can cause the node to panic.
Workaround: Let the node reboot after the panic.
PSCALE-261976 MPU Upload Complete can fail with Malformed XML some AWS SDK like Java SDK. The parse of the
XML body had issues with the XML Declaration in the body.
Workaround: Some S3 SDK will work like the Boto3, just the Java SDK have issues in MPU Complete.
PSCALE-230379 If a node is flooded with many RDMA connection requests, any transient error may prevent the node
from continuing to negotiate connection requestions. This presents itself as a timeout on mount when
RDMA is used: # mount -t nfs -o proto=rdma,port=20049 [Link]:/ifs/
data /mnt/psc
The server still responds to TCP mounts, and the nfs server process is still active: # mount -t nfs
-o proto=tcp,port=2049 [Link]:/ifs/data
# (success)
This does not trigger a failover event, since existing RDMA mounts and TCP, UDP mounts continues
to be served.
6
Workaround: You may attempt a restart of the nfs process using a likewise or using a process
signal: /usr/likewise/bin/lwsm restart nfs
The above command clears the RDMA connection request queue. However, there is the potential for
the connection request queue to be populated before the service starts up again if there is a flood
of RDMA traffic against the node, e.g automount. Dell Technologies recommends disabling automount
until the process has successfully restarted and is accepting new RDMA connections.
Where to get help
The Dell Technologies Support site contains important information about products and services including drivers, installation
packages, product documentation, knowledge base articles, and advisories.
A valid support contract and account might be required to access all the available information about a specific Dell Technologies
product or service.
Additional options for getting help
This section contains resources for getting answers to questions about PowerScale products.
Dell Technologies Support ● Contact Technical Support
Telephone support ● United States: 1-800-SVC-4EMC (1-800-782-4362)
● Canada: 1-800-543-4782
● Worldwide: 1-312-725-5401
● Local phone numbers for a specific country or region are available at Contact
Technical Support.
PowerScale OneFS Documentation ● PowerScale OneFS Info Hubs
Info Hubs
Dell Knowledge Base (KB) articles KB articles are available on the Dell Technologies support site.
Installation and upgrade considerations
This section includes resources to consider when installing and upgrading OneFS.
OneFS information and documentation
You can access technical information about Dell Technologies PowerScale products, OneFS product documentation, and other
resources through the following channels.
Table 2. More PowerScale resources
Channel Description
PowerScale Info Hubs PowerScale info hubs organize PowerScale documentation into topic areas, making it
simple to find content.
PowerScale OneFS Product Page The Dell Technologies PowerScale OneFS product page provides access to PowerScale
product documentation and current software releases. Access to this page requires that
you log in to the Dell Technologies Online Support site.
PowerScale OneFS Patches For the most up-to-date list of patches that are available for the version of OneFS
running on your cluster, see the Current Patches article on the Dell Technologies Online
Support site.
Dell Advisories (DSA, DTA) A Dell Technologies security advisory (DSA) is published when a Dell Technologies
product is impacted by a security vulnerability. A DSA provides a fix that customers must
apply.
7
Table 2. More PowerScale resources (continued)
Channel Description
A Dell Technologies technical advisory (DTA) is published when Dell Technologies
identifies an issue that may cause serious negative impact to a production environment.
For the most up-to-date list of PowerScale Advisories, see Dell Advisories for PowerScale
OneFS. You can also subscribe to receive advisories automatically. See article 000018135
for steps to subscribe.
Healthcheck Info Hub The isi healthcheck command enables you to evaluate the status of specific
software and hardware components of the cluster and the cluster environment. See the
Dell Technologies PowerScale Healthcheck Info Hub for details.
Online Help Select Online Help from the Help menu in the OneFS web administration interface.
8
Notes, cautions, and warnings
NOTE: A NOTE indicates important information that helps you make better use of your product.
CAUTION: A CAUTION indicates either potential damage to hardware or loss of data and tells you how to avoid the
problem.
WARNING: A WARNING indicates a potential for property damage, personal injury, or death.
Copyright © 2016 - 2025 Dell Inc. All Rights Reserved. Dell Technologies, Dell, and other trademarks are trademarks of Dell Inc. or its
subsidiaries. Other trademarks may be trademarks of their respective owners.
