Planned Network & UFM Maintenance AP-IN region
Runpod · 最近儲存的提供方狀態：scheduled
官方來源文字以原始語言顯示。
最近儲存的事件狀態不是當前服務狀態。來源列表可能不完整，事件從來源消失不確認恢復。請開啟當前報告或官方來源檢視較新證據。
已儲存事件詳情
- 提供方狀態
- scheduled
- 提供方影響程度
- maintenance
- 提供方記錄建立時間
- 24 10月 2026, 18:30:00 UTC
- 提供方記錄更新時間
- 24 10月 2026, 18:30:00 UTC
- 計劃視窗
- 24 10月 2026, 18:30:00 UTC → 24 Oct 2026, 21:30:00 UTC
提供方報告的受影響元件
此已儲存修訂未提供元件關聯，或尚未恢復這些關聯。
這些關聯描述此事件的報告範圍。不能確認當前元件可用性或已核實依賴關係。
記錄建立時間未必是故障開始時間。未報告的時間保持不可用。我們不根據收集時間計算故障時長。
Runpod 公佈的平臺資源及當前或未來報告。這些是提供方報告，不是對任何模型、賬戶或地區的獨立探測。
已儲存修訂中的提供方更新
最新優先。最多顯示最近 20 份已儲存內容修訂中的 100 條不同更新。同一提供方時間的文字修訂單獨保留。
- published update
Please be advised that scheduled network maintenance will take place during the following window. This maintenance involves two separate, independent activities grouped together to minimize the overall number of maintenance events. Maintenance Window PDT: October 24, 2026 | 11:30 AM – 2:30 PM IST: October 25, 2026 | 12:00 AM – 3:00 AM Scope: All machines in AP-IN region Planned Activities & Expected Impact 1. DC Internet Traffic Optimization Impact: Users may experience minor network fluctuations, jitter, or latency for approximately 15 minutes during the maintenance window. 2. Enable SHARP on UFM Impact: Running GPU jobs may be impacted or interrupted for up to 120 minutes during this activity. Unaffected Services: H100 accessibility and standard Ethernet/CPU workloads will remain fully operational and unaffected. User Action Required GPU Workloads: Please be aware that active GPU jobs may be interrupted during this maintenance window. In the event of a GPU workload failure or interruption, the affected job will need to be re-initiated by the respective user once the activity is complete. General Network: No action is required for the Internet Traffic Optimization, but please plan accordingly for the brief period of expected network latency.
在 21 10月 2026, 16:12:53 UTC 的已儲存修訂中觀察到