> For the complete documentation index, see [llms.txt](https://docs.silk.us/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.silk.us/releases/2026/v9.5.20/fixes.md).

# Fixes

Generally available since May 6, 2026, with the latest refresh on October 20, 2026. Review the critical, highly recommended, and recommended fixes included in Silk v9.5.20.

Fixes from the October 2026 refresh are listed first, followed by the fixes included in the initial v9.5.20 release.

## October 2026 refresh fixes

These fixes are available since October 20, 2026.

### Critical fixes

<table><thead><tr><th width="116.2962646484375">Fix ID</th><th>Description</th></tr></thead><tbody><tr><td>153464<br>150650</td><td>Enhanced management takeover coordination and connectivity validation to preserve DataPod availability when a c.node running the SMC becomes network-isolated.<br><br>This fix addresses a scenario where a takeover could cause a PMC crash, resulting in a DataPod stall and loss of host I/O access.</td></tr><tr><td>152917</td><td>Improved handling of host requests rejected due to insufficient resources so they receive a response rather than remain pending.<br><br>This prevents outstanding commands from blocking back-end reconfiguration during c.node phase-in, which could cause the DataPod to go offline and interrupt host I/O.</td></tr><tr><td>150881</td><td>Resolved an issue where an upgrade could automatically re-enable disabled Remote Metadata Read (RMDR) parameters, causing inconsistent settings that could stall the DataPod and interrupt host I/O.<br><br>RMDR settings now persist through the upgrade process.</td></tr><tr><td>150293</td><td>Fixed an issue where a c.node becoming stuck in the <code>set_disk_state</code> operation could cause the DataPod to enter a degraded state.<br><br>The system now initiates takeover of the affected c.node rather than entering a degraded state.</td></tr><tr><td>149504</td><td>Resolved a race condition during volume unmapping where completed reset operations weren’t tracked correctly, blocking subsequent provisioning and management operations.<br><br>This fix prevents this condition from causing the PMC to enter a degraded state.</td></tr></tbody></table>

### Recommended fixes

<table><thead><tr><th width="115.555419921875">Fix ID</th><th>Description</th></tr></thead><tbody><tr><td>150991</td><td>Corrected an issue that caused performance spikes to appear in the GUI analysis screen.</td></tr><tr><td>150844</td><td>Corrected AWS service quota handling to preserve applied quota values instead of overwriting them with AWS defaults. This prevents false quota-exceeded errors that could block DataPod creation.</td></tr><tr><td>150559</td><td>Improved Azure Blue-Green upgrade handling to ensure the new Flex IP address is updated on c.nodes, even if an earlier preparation stage failed.<br><br>Corrected source IP handling in Flex to preserve the caller’s original IP address, allowing authorized requests from the Flex SSH script to be validated.</td></tr></tbody></table>

## Critical fixes

Fixes for issues that could significantly affect system availability, data security, or the ability to perform essential operations. Prioritize applying these fixes.

<table><thead><tr><th width="116.2962646484375">Fix ID</th><th>Description</th></tr></thead><tbody><tr><td>Cases:<br>136401<br>124706</td><td>Fixed an issue where connectivity interruptions prevented an NDU workflow from completing as expected.</td></tr><tr><td>149277</td><td>Fixed an issue where volume resize operations could stall if the volume was mapped to a host definition that no longer existed or was otherwise invalid.<br><br>This update improves resiliency during volume management operations and prevents unexpected system stalls caused by stale host mappings.</td></tr><tr><td>142965</td><td>Improved coordination between system recovery and takeover mechanisms to help prevent unintended c.node restart scenarios during recovery operations.<br><br>Enhanced recovery timing and decision handling improves platform stability and reduces the risk of system-wide disruption during complex recovery events.</td></tr><tr><td>142830</td><td>Optimized Data Manager memory utilization during recovery and monitoring operations to reduce excessive memory consumption under high recovery load conditions.<br><br>Improved recovery efficiency and resource handling to help maintain platform stability and reduce risk of system disruption during complex recovery scenarios.</td></tr><tr><td>140080</td><td><p>Improved handling of unexpected disk detach events and recovery scenarios that led to stalled recovery operations and degraded host I/O access.</p><p><br>Validation and recovery logic for Azure Pv2 disk state changes is enhanced to help prevent recovery interruptions caused by inconsistent detach signaling during failover and recovery workflows.</p></td></tr><tr><td>148000</td><td><p>Resolved an issue where an NDU failed during Management Controller (MC) updates, causing a system stall and loss of host I/O access during upgrade operations.</p><p><br>Improved handling of MC failover operations now supports more reliable upgrade execution and reduces the risk of disruption during NDUs.</p></td></tr></tbody></table>

## Highly recommended fixes

Fixes for significant functional or operational issues that aren’t expected to affect overall system availability or security. Applying these fixes is strongly recommended to maintain reliable operation and avoid potentially disruptive behavior.

<table data-search="false"><thead><tr><th width="114.8148193359375">Fix ID</th><th>Description</th></tr></thead><tbody><tr><td>149274</td><td>Fixed an issue where TRIM/UNMAP commands weren’t processed correctly for Windows file systems using a 64K allocation unit size, preventing expected space reclamation and causing inaccurate physical capacity utilization reporting.</td></tr><tr><td>148283</td><td><p>Fixed an issue in Windows cluster environments where attempting to re-register an existing key could trigger an unexpected LUN reset event.</p><p>This update improves host registration stability and helps prevent unintended disruption during cluster operations.</p></td></tr><tr><td>148216</td><td><p>Improved recovery handling for CHAP authentication-related scenarios that caused a system to enter a degraded management state during recovery operations.</p><p><br>Enhanced validation and recovery coordination logic improves system stability and reduces the risk of operational disruption during authentication and failover workflows.</p></td></tr><tr><td>147474</td><td><p>Resolved an issue where stale address allocation information left after m.node decommissioning interfered with subsequent scale-up operations.</p><p><br>Improved cleanup and validation handling during m.node lifecycle operations to help ensure more reliable scale-up execution and reduce risk of address allocation conflicts when adding new m.nodes.</p></td></tr><tr><td>147127</td><td>Resolved an issue where Flex could display or validate incorrect m.node VM SKUs during DataPod expansion and m.node addition workflows.<br><br>Improved SKU validation and synchronization handling ensures supported VM types are presented correctly during deployment and scale-up operations.</td></tr><tr><td>146102</td><td><p>Improved detection and handling of c.node OS disk availability issues that impacted recovery and failover operations.</p><p><br>Added enhanced monitoring and event reporting to provide greater visibility into OS disk state changes and accelerate troubleshooting during infrastructure recovery scenarios.</p></td></tr><tr><td>145838</td><td><p>Resolved an issue where LDAP configuration settings became unavailable following MC failover and recovery events.</p><p><br>Improved persistence and recovery handling for LDAP configuration data helps maintain authentication connectivity and reduces the risk of configuration loss during takeover operations.</p></td></tr></tbody></table>

## Recommended fixes

Fixes for smaller or more limited issues that may affect specific workflows, behaviors, or user experiences. These fixes are recommended for customers who use the affected functionality.

<table><thead><tr><th width="115.555419921875">Fix ID</th><th>Description</th></tr></thead><tbody><tr><td>148213</td><td><p>Resolved an issue where backend storage journal conditions triggered unnecessary low-resource operational alerts.</p><p><br>Corrected journal sizing behavior associated with this condition to restore normal backend operation handling.</p></td></tr><tr><td>146431</td><td><p>Resolved an issue where exporting the Replication RPO History report generated an empty output file despite displaying data in the report view.</p><p><br>Corrected CSV export handling to ensure replication RPO history data is exported successfully from Clarity reports.</p></td></tr><tr><td>138739</td><td><p>Resolved an issue where cluster administrator passwords containing certain special characters weren’t passed correctly to the DataPod during deployment and configuration workflows.</p><p><br>Improved password handling and validation behavior ensures supported special characters are processed consistently across deployment operations.</p></td></tr></tbody></table>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.silk.us/releases/2026/v9.5.20/fixes.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
