Active incident · posted by the vendor
View on source(opens in a new tab)INC20000150
- Identified
“Current status: Since our last update, we have seen periods of improved provisioning performance while preserving available compute resources and coordinating closely with our third-party cloud provider. Resource availability remains constrained and variable, and provisioning delays or failures may still occur. The provider continues to project improved resource availability in the specified region by the end of October. We will provide another update within 7 days, or sooner if there is a material change. Customer experience: Customers in the specified region may still encounter delays, timeouts, or failures when running Snowpark Container Services workloads, including starting or resuming Notebooks, provisioning or resuming compute pools, and setting up OpenFlow runtimes. Virtual Warehouses: Customers using second generation (Gen2) warehouses may experience delays or intermittent failures when starting or resuming their warehouses, which can extend query run times or delay completion of data pipelines. Warehouses that are already running are generally less affected. ETA: Based on information from our third-party cloud platform provider, we currently expect additional compute resources to become available in the specified region by late October, although the benefit may vary by workload. We'll provide an updated ETA if that timeline changes. Workaround: While we remain focused on reducing provisioning times and building on measures that have already shown improvement, we recommend the following actions, where feasible for affected customers. -Delay migrating warehouse-based Notebooks to SPCS where possible. -For customers experiencing Gen2 warehouse start or resume delays, and where appropriate for their workload, moving the affected warehouse to Gen1 may provide short-term relief. Gen1 warehouses are not affected by this capacity constraint, but Gen1 capacity is limited and performance characteristics may differ. -Contact Snowflake Support for case-by-case assistance with SPCS compute-pool failures. Incident start time: 08:41 UTC August 09, 2026 Preliminary root cause: A third-party cloud platform is experiencing constrained compute capacity in the specified region, limiting the availability of compute resources for Snowpark Container Services and causing delays or failures when provisioning or resuming virtual warehouses.”
- Identified
“Current status: We've received an update from our third-party cloud provider that improvement for resource availability in this region is now projected by the end of October. In the meantime, we’re continuing our efforts to reduce virtual warehouse provisioning times, building on measures that have already shown improvement. In parallel, we’re preserving available compute resources and continuing to closely coordinate with our third-party cloud provider on resource availability updates in the specified region. Resource availability continues to vary with demand, and provisioning delays or failures may still occur. We’ll provide another update within 7 days. Customer experience: Over the past few weeks, we've seen periods of fewer failures as we continue working to preserve available compute resources. However, resource availability at our third-party cloud provider remains constrained. Customers in the specified region may still encounter failures, timeouts, or delays when running Snowpark Container Services workloads, including starting or resuming Notebooks and provisioning or resuming compute pools. Creating additional workloads in new accounts adds to demand for these limited resources. Customers setting up Snowpark Container Services services or jobs, or Openflow runtimes, in those accounts may therefore encounter more frequent provisioning failures or be unable to complete setup. Customers using second generation (Gen2) warehouses may also continue to experience delays or intermittent failures when starting or resuming their warehouses, which can extend query run times or delay completion of data pipelines. ETA: Our third-party cloud platform provider remains committed to making additional compute resources available, which are now projected to be ready by late-October. We'll provide an updated ETA if that timeline changes. Workaround: Customers running Notebooks on virtual warehouses can delay migrating to Snowpark Container Services until this issue is resolved. Customers experiencing virtual warehouse resume delays can migrate affected warehouses to first generation (Gen1) using ALTER WAREHOUSE <warehouse_name> SET GENERATION = '1';. Gen1 warehouses are not affected by this capacity constraint. This is a short-term mitigation; Gen1 capacity is limited and performance characteristics may differ. There are currently no general workarounds available for Snowpark Container Services compute pool impact. Customers with specific compute pool failures should contact Snowflake Support for case-by-case assistance. Incident start time: 08:41 UTC August 09, 2026 Preliminary root cause: A third-party cloud platform experienced constrained compute capacity, causing limited availability of compute resources for Snowpark Container Services and delays or failures when provisioning or resuming virtual warehouses in the specified region.”
- Identified
“Current status: We've taken additional steps with our warehouse team to reduce provisioning latencies for virtual warehouses, and we've observed some improvement in provisioning latencies as a result. Resource availability remains variable and correlated to overall demand, with some intermittent delays occurring during high-volume times. We're continuing to work closely with our warehouse team to maintain available capacity and keep provisioning times low. Our third-party cloud platform provider has indicated that the additional compute capacity remains on track, and we're continuing to coordinate with them to restore the needed capacity. We'll provide another update within 7 days. Customer experience: Customers hosted in the specified region may intermittently be unable to use Snowpark Container Services. Affected customers may experience failures, timeouts, or delays when running Snowpark Container Services workloads, such as starting or resuming Notebooks, or provisioning or resuming compute pools. Notebooks running on virtual warehouses are not affected. Additionally, starting from 04:00 UTC September 04, 2026, customers hosted in the specified region with workloads on second generation (Gen2) warehouses may experience delays or intermittent failures when resuming or starting virtual warehouses. This may result in longer than expected query run times or delays in completing data pipelines. Warehouses that are already running are largely unaffected. ETA: Based on information from our third-party cloud platform provider, we expect additional compute resources to become available in the specified region by mid-October. We'll provide an updated ETA as we receive more information. Workaround: While we're focused on remediation, customers running Notebooks on virtual warehouses can delay migrating to Snowpark Container Services until this issue is resolved. For customers experiencing impact to virtual warehouses, and who have already performed a failover, we recommend remaining in the alternate region until we have received additional capacity allocation from the third-party provider. Customers who are continuing to experience impact to virtual warehouses may see additional relief by migrating workloads to first generation (Gen1) warehouses, which remain less impacted by this issue. Incident start time: 08:41 UTC August 09, 2026 Preliminary root cause: A third-party cloud platform provider is experiencing constrained compute capacity, affecting compute resource provisioning, and causing limited availability of compute resources as well as delays and intermittent failures when resuming or starting virtual warehouses in the specified region.”
- Identified
“Current status: Our third-party cloud platform provider has informed us that they expect to be able to provide additional capacity allocation by mid-October. In the meantime, we are working to preserve compute resources available to us, reduce non-essential resource consumption, and investigate any alternative methods to further reduce impact. Resource availability remains variable and correlated to overall demand, with some intermittent delays occurring during high-volume times. We'll provide another update by 23:00 UTC September 17, 2026. Customer experience: Customers hosted in the specified region may intermittently be unable to use Snowpark Container Services. Affected customers may experience failures, timeouts, or delays when running Snowpark Container Services workloads, such as starting or resuming Notebooks, or provisioning or resuming compute pools. Notebooks running on virtual warehouses are not affected. Additionally, starting from 04:00 UTC September 04, 2026, customers hosted in the specified region with workloads on second generation (Gen2) warehouses may experience delays or intermittent failures when resuming or starting virtual warehouses. This may result in longer than expected query run times or delays in completing data pipelines. Warehouses that are already running are largely unaffected. ETA: Based on information from our third-party cloud platform provider, we anticipate a solution in mid-October. Workaround: While we're focused on remediation, customers running Notebooks on virtual warehouses can delay migrating to Snowpark Container Services until this issue is resolved. For customers experiencing impact to virtual warehouses, and who have already performed a failover, we recommend remaining in the alternate region until we have received additional capacity allocation from the third-party provider. Customers who are continuing to experience impact to virtual warehouses may see additional relief by migrating workloads to first generation (Gen1) warehouses, which remain less impacted by this issue. Incident start time: 08:41 UTC August 09, 2026 Preliminary root cause: A third-party cloud platform provider is experiencing constrained compute capacity, affecting compute resource provisioning, and causing limited availability of compute resources as well as delays and intermittent failures when resuming or starting virtual warehouses in the specified region.”
- Identified
“Current status: We've made additional adjustments to further reduce non-essential resource consumption and preserve available compute resources in the specified region. We're continuing to develop a software update to enable additional compute resources, and our other mitigation actions remain in place. We're also continuing to coordinate with our third-party cloud platform provider to restore the compute capacity needed for Snowpark Container Services. We'll provide another update within 7 days. Customer experience: Customers hosted in the specified region may intermittently be unable to use Snowpark Container Services. Affected customers may experience failures, timeouts, or delays when running Snowpark Container Services workloads, such as starting or resuming Notebooks, or provisioning or resuming compute pools. Notebooks running on virtual warehouses are not affected. ETA: Based on information from our third-party cloud platform provider, we expect this capacity constraint to persist for several more months. We'll provide an updated ETA as we receive more information. Workaround: While we're focused on remediation, customers running Notebooks on virtual warehouses can delay migrating to Snowpark Container Services until this issue is resolved. Incident start time: 08:41 UTC August 09, 2026 Preliminary root cause: A third-party cloud platform experienced constrained compute capacity, which caused limited availability of compute resources in the specified region.”
- 3 earlier updates on the source(opens in a new tab)