Dell PowerStore DSA-2026-479: Patch Critical Storage Flaws

Enterprise storage array protected from a command-injection path during a security update workflow

Dell’s PowerStore security update DSA-2026-479 deserves immediate attention from storage and infrastructure teams. The advisory covers a large set of PowerStore and SDNAS weaknesses, including unauthenticated attack paths scored 9.8 and authenticated paths that can end in root-level command execution. Dell released the bulletin on October 30, 2026; individual CVE records, including CVE-2026-103813, became public on November 13.

The operational answer is straightforward: identify PowerStore T and Q appliances on the affected 5.1.x branch, restrict management and NAS exposure, run Dell’s current health checks, and move to PowerStoreOS 5.1.0.0 build 2798105 or later. This guide explains why the update is urgent, how to prioritize clusters, and how to validate the work without running exploit code.

What Dell disclosed in DSA-2026-479

Dell advisory DSA-2026-479 groups weaknesses across the PowerStore management plane and the SDNAS data-services layer. The most serious entries include:

  • CVE-2026-116443, a stack-based buffer overflow in the NAS TFTP service. Dell says an unauthenticated remote attacker could cause denial of service or potentially execute code. CVSS 3.1: 9.8.
  • CVE-2026-116440, an SDNAS buffer overflow reachable by an unauthenticated remote attacker, with potential denial of service or code execution. CVSS 3.1: 9.8.
  • CVE-2026-116448 and CVE-2026-116449, OS command-injection flaws that can allow an attacker with internal network access to run arbitrary commands as root on a NAS node. CVSS 3.1: 9.8.
  • CVE-2026-116439, an unauthenticated path-traversal issue that can permit arbitrary file reading, modification, or deletion on a NAS node. CVSS 3.1: 9.8.
  • CVE-2026-103813, a network-reachable command-injection flaw through which a limited authenticated user could escalate to root. CVSS 3.1: 9.1.
  • CVE-2026-111545, SQL injection that Dell says can expose stored private keys and allow trusted-certificate tampering. CVSS 3.1: 8.8.

This is not one bug with one prerequisite. The advisory combines unauthenticated, adjacent-network, internal-network, authenticated, and local attack conditions. A cluster with no internet-facing management interface can still matter because compromised internal hosts, stolen appliance credentials, backup networks, and NAS protocols can supply the required position.

Affected and remediated versions

Dell’s remediation table covers PowerStore 500T, 1000T, 1200T, 1500, 3000T, 3200Q, 3200T, 5000T, 5200Q, 5200T, 5500, 7000T, 9000T, 9200T, and 9500 systems on PowerStoreT OS 5.1.x.

Branch in advisoryStatusRequired action
PowerStoreT OS 5.1.x before 5.1.0.0-2798105AffectedPlan upgrade to 5.1.0.0-2798105 or later
PowerStoreT OS 5.1.0.0-2798105 or laterRemediated by DSA-2026-479Verify build, health, and exposure controls
Any branch not explicitly represented in the advisory tableReview requiredUse the service tag and Dell support guidance; do not infer safety from a version mismatch

Dell’s PowerStoreOS matrix lists 5.1.0.0 build 2798105 with an October 29 release date. Do not compare only the four dotted components. The build number is part of the fixed version, so record the complete string during inventory.

Why storage teams should treat this as a control-plane incident risk

Root access changes the blast radius

Several flaws can reach root or system privileges. On a storage appliance, that is not equivalent to compromising an ordinary application server. The appliance mediates volumes, file services, snapshots, replication, backup workflows, certificates, and administrative identities. A successful attacker may be positioned to disrupt availability, alter trusted relationships, or interfere with recovery data.

Low-privilege accounts are part of the attack surface

CVE-2026-103813 is important even though its CVSS vector requires high privileges. Dell’s description says a limited authenticated user can potentially reach root. Review local accounts, directory-backed roles, service accounts, dormant operators, API users, and support access. A “read-only” or narrowly scoped account should not be treated as harmless while an affected build remains in service.

NAS exposure may differ from management exposure

Some critical SDNAS entries are reachable without authentication or from an internal network. Map the actual interfaces and routed paths for TFTP, NFS, NDMP, SSH/SFTP, and management services. A firewall rule protecting PowerStore Manager does not automatically constrain every NAS protocol or storage-network peer.

Immediate containment and prioritization

  1. Inventory every cluster. Record model, full PowerStoreOS version and build, service tag, NAS services, management interfaces, authentication source, business owner, and maintenance window.
  2. Prioritize reachable systems. Put clusters with broadly reachable NAS services, shared administrative networks, third-party support paths, or recently changed accounts at the top of the queue.
  3. Reduce access before maintenance. Limit management interfaces to dedicated administrator networks or a controlled jump path. Restrict NAS protocols to required clients and networks. Do not make emergency ACL changes without validating host and replication dependencies.
  4. Review identities. Disable stale accounts, rotate credentials where compromise is plausible, verify directory-group membership, and preserve an approved break-glass path.
  5. Preserve evidence. Export relevant alerts, audit events, account changes, authentication records, configuration changes, and network telemetry before logs roll over. If suspicious activity exists, coordinate with incident response before rebooting or upgrading.

If you discover suspicious sessions or configuration changes, use an evidence-first process similar to our Ivanti Connect Secure incident-response guide. For a broader model of exposure-driven prioritization, see the VeloCloud Orchestrator response guide.

Hands-on lab: classify a PowerStore inventory safely

This lab is an offline triage helper, not a vulnerability scanner. It does not connect to a PowerStore appliance, attempt authentication, or send exploit traffic. Use an isolated administration workstation with Python 3 and a CSV created from approved inventory data.

Prerequisites

  • Python 3.9 or later
  • An inventory export with cluster name, model, full version, and management exposure
  • No secrets, passwords, tokens, or customer data in the CSV

Create the inventory file

cat > powerstore_inventory.csv <<'CSV'
cluster,model,version,management_exposure
prod-san-01,PowerStore 5200T,5.1.0.0-2798000,admin-vlan-only
dr-san-01,PowerStore 1200T,5.1.0.0-2798105,vpn-admin-vlan
lab-san-01,PowerStore 500T,4.3.1.2-2771239,isolated-lab
CSV

Run the classifier

#!/usr/bin/env python3
import csv, re, sys

FIXED = (5, 1, 0, 0, 2798105)
VERSION_RE = re.compile(r"^(d+).(d+).(d+).(d+)-(d+)$")

def classify(version):
    match = VERSION_RE.fullmatch(version.strip())
    if not match:
        return "REVIEW", "invalid or missing version format"
    parsed = tuple(int(part) for part in match.groups())
    if parsed[:2] != (5, 1):
        return "REVIEW", "outside the 5.1.x branch listed in the advisory"
    if parsed < FIXED:
        return "AFFECTED", "below 5.1.0.0-2798105"
    return "REMEDIATED", "at or above 5.1.0.0-2798105"

with open(sys.argv[1], newline="", encoding="utf-8") as handle:
    for row in csv.DictReader(handle):
        status, reason = classify(row["version"])
        print(f'{row["cluster"]}: {status} | {row["version"]} | {reason}')
python3 powerstore_dsa_2026_479_audit.py powerstore_inventory.csv

Expected output

prod-san-01: AFFECTED | 5.1.0.0-2798000 | below 5.1.0.0-2798105
dr-san-01: REMEDIATED | 5.1.0.0-2798105 | at or above 5.1.0.0-2798105
lab-san-01: REVIEW | 4.3.1.2-2771239 | outside the 5.1.x branch listed in the advisory

Interpretation: “REVIEW” is deliberate. DSA-2026-479’s table identifies a 5.1.x remediation threshold; the helper does not declare another branch safe or vulnerable. Confirm non-5.1 systems against their product support page and Dell guidance.

Troubleshooting and cleanup

  • If a row reports an invalid version, copy the full PowerStoreOS version and build instead of a shortened value such as 5.1.
  • If a required column is missing, preserve the header names exactly.
  • When triage is complete, move the report into the approved change record or delete the temporary CSV according to your data-handling policy.

Plan the upgrade without skipping health checks

Dell documents a non-disruptive upgrade path, but “non-disruptive” still depends on cluster and host readiness. The PowerStore pre-upgrade health-check procedure recommends installing the latest compatible Health Check package, running preliminary checks early enough to resolve problems, and running the relevant checks again immediately before the upgrade.

Use this change sequence:

  1. Confirm the target package for the exact model and current release, and review the associated release notes.
  2. Verify host multipathing, path redundancy, replication state, capacity, active alerts, and support connectivity.
  3. Install the latest compatible Health Check package and run System Check plus the Pre-Upgrade Health Check.
  4. Resolve every blocking error. Escalate unclear warnings to Dell support instead of suppressing them.
  5. Export the current configuration and evidence needed by your recovery and audit procedures.
  6. Perform the upgrade during an approved window, with application owners monitoring latency and path failover.
  7. Verify the complete version string, cluster health, volumes, file services, replication, alerts, authentication, certificates, and client paths.

Dell’s NDU preparation guidance specifically calls for current health-check packages and review of outstanding alerts. Treat those checks as change controls, not optional ceremony.

Detection and post-upgrade review

The public advisory describes impacts, not universal indicators of compromise. Avoid building detections around invented paths or process names. Instead, establish a baseline and review evidence available in your environment:

  • Unexpected local or directory-backed account creation, role changes, password resets, or authentication-source changes
  • Administrative access from new source addresses, unusual jump hosts, or outside approved change windows
  • Changes to NAS services, exports, certificates, replication, remote support, or network configuration
  • Repeated service failures, node restarts, health alerts, or unexplained file-service interruptions
  • Firewall and flow records showing new management or NAS protocol reachability

If evidence suggests exploitation, isolate carefully. Storage systems have many dependencies, and an abrupt network block can affect production or replication. Preserve logs, document the current state, involve storage engineering and incident response, and use Dell support for appliance-level forensics.

Conclusion

DSA-2026-479 is a broad PowerStore security update with multiple credible paths to severe impact. The most useful response is not indiscriminate scanning; it is disciplined asset identification, exposure reduction, identity review, evidence preservation, Dell health checks, and a controlled move to PowerStoreOS 5.1.0.0-2798105 or later.

Start with the clusters that combine affected builds with reachable NAS or management services. Then verify the full fleet, including systems on other release branches that require vendor-specific review. A storage platform is part of the recovery plane, so patching it safely is both a vulnerability-management task and a resilience exercise.

References

Similar Posts