US-LAX-A Storage Incident and Service Restoration Notice

Incident Time: August 2, 2026, 00:51 Hong Kong Time
Affected Services: Selected cloud servers in the US-LAX-A Los Angeles region
Current Status: All affected services have been restored

Dear Customers,

At 00:51 Hong Kong Time on August 2, 2026, a storage hardware link failure occurred on one of the host servers in our US-LAX-A Los Angeles region. As a result, several cloud servers on the affected host were unable to read from or write to their disks, causing a service interruption.
Immediately after the incident was detected, we arranged for remote-hands engineers to conduct an on-site inspection. The investigation confirmed that the issue was caused by a faulty drive-bay connector in the Dell server, preventing part of the original Intel P4500 RAID array from being properly detected and accessed.

As the required replacement component was not available on site, and the vendor’s inspection, RMA and replacement process may take a considerable amount of time, the affected host could not be repaired within an acceptable restoration window. To restore customer services as quickly as possible and avoid a prolonged interruption, we deployed a new, higher-specification host and completed the restoration, rebuilding and migration of all affected instances.

All affected services have now been restored.

Data Status

The incident affected the data state of some instances:
·       Most affected instances were restored using an available data copy dated July 29, 2026;
·       Instances belonging to six customers had no historical copy immediately available and were recreated with their original product specifications and network configurations. These customers will need to redeploy their applications and business data;
·       Data written after the available recovery point remains on the original storage devices but cannot currently be accessed due to the hardware link failure. Further processing will require replacement parts or completion of the vendor’s inspection and RMA process, and we are currently unable to confirm the required timeframe;
·       IP addresses, bandwidth allocations and product specifications remain unchanged. No changes to DNS records, IP allowlists or third-party integrations are required.
Customers who continue to experience issues with system startup, storage, networking or application deployment may submit a support ticket through the customer portal. Related requests will be handled with priority.

Infrastructure Upgrades

As part of the restoration process, we also upgraded the host and network infrastructure serving the LAX-A region:
·       The computing platform was upgraded from Intel Xeon Gold Skylake to AMD EPYC 7003 Series;
·       The storage platform was upgraded from an Intel P4500 RAID array to an Intel P5500 enterprise SSD RAID array;
·       The host network uplink was upgraded to 100Gbps;
·       The LAX-A region now provides AS4837 connectivity for traffic towards Mainland China;
·       BGP Session services are now available in the LAX-A region.
We remain committed to using enterprise-grade server, storage and network platforms. Through hardware redundancy, proactive monitoring and structured incident-response procedures, we make every reasonable effort to maintain the reliability and stability of our services.
However, no hardware or infrastructure architecture can completely eliminate the risk of an unexpected failure. Customers are encouraged to maintain independent and regularly verified backups of databases, website files, application configurations and other critical data. Where business continuity is essential, backup copies should be stored on separate hosts, in another region or on an independent storage platform.
We will continue strengthening our critical spare-parts inventory, storage-link monitoring, off-host backup coverage and disaster-recovery procedures to further improve service continuity and data protection.

Compensation

We sincerely apologize for the service interruption, changes to instance data states and the recovery work caused by this incident.
As compensation:
All existing customers in the LAX-A region will receive a 30-day service extension.

The extension will be applied automatically, and no application or support ticket is required. Please refer to the service expiration date shown in the customer portal for confirmation.

Although the immediate cause of the incident was a hardware connector failure, we recognize that improvements are still required in our spare-parts readiness, backup coverage and recovery procedures. We will take responsibility for the issues exposed by this event and ensure that the necessary corrective measures are implemented.

We once again extend our sincere apologies to all customers in the LAX-A region for the impact caused by this incident.

Thank you for your understanding, trust and continued support.

ISIF
August 3, 2026