Hyperstack Outage History

We are performing an emergency maintenance on the Hyperstack platform and API Services to address a platform issue. During this window, the Hyperstack console and public API may be unavailable or return errors. Existing running workloads (VMs, containers, and Kubernetes clusters) are not affected and continue to run; only management-plane operations are impacted. We expect our services to be restored soon. We will post an update every 60 minutes or sooner if the situation changes. We apologise for the short notice and any disruption caused.

We are continuing to investigate this issue.

As previously communicated via email we are currently in the process of transitioning the Hyperstack Support Solution. We will be temporarily disabling the Live Chat option but we will still be able to respond to queries over email via [email protected]. We apologies for any inconvenience caused by this.

We are currently investigating an issue in the Canada-1 region following emergency maintenance. This has resulted in some hosts not becoming active following reboot. This has also resulted in a reduced selection of deployments available. We are working to resolve this issue as soon as possible.

We are investigating degraded API availability affecting the Hyperstack Console. Some API-dependent workflows may be degraded or unavailable.

We are investigating degraded API availability affecting the Hyperstack Console. Some API-dependent workflows may be degraded or unavailable.

Some customers may be unable to access the Hyperstack console or related services. We are working to restore access as quickly as possible.

Our monitoring has identified network issues that are affecting these two regions. This is currently under investigation and work is underway by our team to return service to expected levels.

We are continuing to investigate this issue.

Summary We are currently investigating an issue affecting cluster reconciliation. The NVIDIA Helm chart repository (https://helm.ngc.nvidia.com/nvidia) is returning errors and appears to be unreachable. Impact Clusters that depend on NVIDIA Helm charts may fail to reconcile to their desired state. Observed Error 502 Bad Gateway when attempting to fetch index.yaml from the NVIDIA Helm repository. Current Status Investigation is ongoing. We are monitoring the upstream NVIDIA repository and assessing potential mitigation options. Next Update We will provide an update as soon as additional information or mitigation steps become available.

We are continuing to investigate this issue.

We are currently investigating a network connectivity issue affecting customer VM reachability from the public internet in CANADA-1 (CA-1) Region. Our on-site team is actively engaged and working to resolve the fault. What we know: • The issue has been isolated to the network layer at the data centre level at CANADA-1 Region only • HyperStack UI Console is impacted, but all the API Services are operational • Internal control plane communication is unaffected, so your VMs are running normally, but external connectivity is impacted • VM-to-VM traffic within the platform is expected to be unaffected • Public internet traffic is unable to reach customer VMs beyond the default gateway We are working directly with our data centre team to restore full connectivity. A further update will be posted within once we have the fix implemented.

We are currently aware of a network-related issue affecting our CANADA-2 Data Centre. Our engineering teams are actively investigating the matter. We will provide further updates on the status and progress as soon as they become available.

We are currently investigating an issue where newly created H100 virtual machines in our CANADA-1 (CA-1) Region may fail to launch and report an error during creation. Other VM/Flavor types and regions are not impacted. Our engineering teams are actively working to identify the root cause and restore normal H100 VM provisioning. We will provide further updates as more information becomes available. If you have any further questions or queries, please contact support via [email protected]

We are currently investigating intermittent failures to launch on-demand Kubernetes clusters in our CA1 region. When failing, the cluster build will respond with an error "Infrastructure reconciliation failed".

Investigating – Connectivity Issues in Canada (CA1) We are currently investigating an issue affecting newly created and restored virtual machines in our Canada (CA1) region, where some instances may not become reachable immediately after launch (e.g. SSH or network access). Our engineering teams are actively working on mitigating the issue and restoring normal service. Existing running instances are not impacted. We will provide further updates as more information becomes available.

We are currently aware of a network-related issue affecting our Canada Data Centre. Our engineering teams are actively investigating the matter. We will provide further updates on the status and progress as soon as they become available.

We are currently aware of a service disruption affecting HyperStack and API Services. This issue appears to be related to a wider incident on the Cloudflare Global Network, which Cloudflare is actively investigating. Our team is monitoring the situation closely and working to mitigate any impact on our platform. We will provide further updates as more information becomes available. Time detected: Nov 18, 2025 – 11:48 UTC

We are currently investigating reports of a service impairment impacting the US-1 region. Customers in this region may experience degraded performance or issues when deploying new resources. Our engineering team is actively working to identify the root cause. We will share further updates as soon as more information becomes available.

We are currently investigating reports of a potential service impairment affecting the US-1 region. Customers in this region may experience degraded services. Our engineering team is actively working to identify the root cause and will provide an update as soon as more information becomes available.