We test Corax against real-world infrastructure failures across every vendor, platform, and scenario. Browse the results below.
The HPE Aruba ClearPass Policy Manager cluster primary node crashes, and the secondary node's database is out of sync, causing all 802.1X authentications to fail across the entire campus network.
An HPE Nimble AF60 all-flash array fails to replicate to the DR site for 12 hours after a WAN circuit issue causes the replication stream to abort repeatedly, creating a growing RPO gap.
An HPE ProLiant DL380 Gen10 Plus triggers a thermal emergency via ILO when the HVAC system fails in the server room, with CPU temperatures approaching the shutdown threshold.
A Dell VxRail E660F node experiences a motherboard failure, removing compute and storage capacity from the vSAN cluster. The cluster enters a reduced state with lower fault tolerance.
A Dell PowerStore 5200 experiences an unexpected controller A failure, triggering automatic failover to controller B. During failover, all I/O is paused for 15 seconds, causing application-level errors.
A Dell EqualLogic PS6210 SAN group member experiences multiple disk failures in the same RAID set, taking a critical iSCSI volume offline and causing all VMs on that datastore to pause.
The iDRAC BMC on a Dell PowerEdge R750 becomes completely unresponsive after a firmware update failure, losing all out-of-band management capability for a critical database server.
All 85 Juniper Mist-managed APs lose connectivity to the Mist cloud dashboard after a rule change blocks the required outbound ports, leaving APs operational but unmanaged.
A Juniper MX480 border router experiences BGP session flapping with two upstream ISPs due to a line card memory error, causing route oscillation and intermittent internet connectivity for the entire organization.
A Juniper EX4300 virtual chassis of 4 members splits into two partitions when a VCP cable is severed during cable management, causing half the access ports to lose their trunk uplinks.
A Juniper SRX4600 chassis cluster experiences a split-brain condition when both fabric links fail simultaneously, causing both nodes to assume the primary role and creating duplicate gateways on the network.
The SonicWall TZ670 SD-WAN policy engine fails to detect WAN link degradation, continuing to route critical VoIP traffic over a congested ISP link instead of failing over to the backup MPLS circuit.
The SonicWall NSsp 13700 running carrier-grade NAT exhausts its NAT translation table, preventing new outbound connections while existing sessions remain active.
The SonicWall NSA 2700 DPI-SSL engine crashes under load, causing all HTTPS traffic to be dropped instead of bypassed, resulting in complete internet outage for 300 users.
A SonicWall TZ670 content filter policy update blocks access to a critical cloud accounting application by miscategorizing it as gambling/gaming, affecting all finance department workflows.
The PA-3260 URL filtering database becomes corrupted during an update, causing all web traffic to be categorized as 'not-resolved' and bypassing URL filtering policies entirely.
The PA-5250 WildFire cloud connection fails, causing all unknown file verdicts to default to 'allow' per policy. Potentially malicious files bypass inspection for 3 hours before detection.
Panorama loses connectivity to all 12 managed PA firewalls after a network change isolates the management VLAN. Firewalls continue forwarding but cannot receive policy updates or log to Panorama.
A Palo Alto threat prevention content update incorrectly classifies ERP traffic as a command-and-control callback pattern, blocking all employee access to the SAP ERP system.
The Palo Alto PA-5250 GlobalProtect gateway stops accepting new VPN connections after a configuration push corrupts the portal agent configuration. 400 remote workers cannot connect.
Every scenario is tested against Corax's Neural Engine in a production environment with AI-powered root cause analysis.
Tests run continuously as new infrastructure patterns are added.