本文列出了功能可用性存在区域差异的功能。
被至少一个区域拒绝的功能包括:
系统表
无服务器
AI 和机器学习
模型服务
注意
下表中不包含某些功能。 这些功能称为“指定服务”,在处理客户内容时使用地理位置来管理数据驻留。 Geo 区域是由一组数据中心区域组成的,Databricks 使用这些区域来提供有关数据处理位置的可预测性和透明度。 请参阅 Databricks 指定服务。
下表列出了 Databricks 支持的 Azure 区域。 某些功能仅在部分区域中受支持,该表指示某个区域是否支持其中每个功能。 如果某个功能在所有区域都受支持,则表中不包含该功能。
注意
Azure 政府区域 usgovaz
和 usgovva
不支持以下任何功能。 此外,这些区域不支持 Databricks SQL 或 Unity 目录。
系统表可用性
区域 | 位置 | 系统表 |
---|---|---|
australiacentral |
澳大利亚中部 | ✓ |
australiacentral2 |
AustraliaCentral2 | ✓ |
australiaeast |
澳大利亚东部 | ✓ |
australiasoutheast |
澳大利亚东南部 | ✓ |
brazilsouth |
巴西南部 | ✓ |
canadacentral |
CanadaCentral | ✓ |
canadaeast |
加拿大东部 | ✓ |
centralindia |
CentralIndia | ✓ |
centralus |
CentralUS | ✓ |
chinaeast2 |
ChinaEast2 | |
chinaeast3 |
ChinaEast3 | |
chinanorth2 |
ChinaNorth2 | |
chinanorth3 |
ChinaNorth3 | |
eastasia |
EastAsia | ✓ |
eastus |
EastUS | ✓ |
eastus2 |
EastUS2 | ✓ |
francecentral |
FranceCentral | ✓ |
germanywestcentral |
德国西中部 | ✓ |
japaneast |
JapanEast | ✓ |
japanwest |
日本西部 | ✓ |
koreacentral |
KoreaCentral | ✓ |
mexicocentral |
墨西哥中心 | ✓ |
northcentralus |
NorthCentralUS | ✓ |
northeurope |
NorthEurope | ✓ |
norwayeast |
挪威东部 | ✓ |
qatarcentral |
卡塔尔中央 | ✓ |
southafricanorth |
南非北 | ✓ |
southcentralus |
SouthCentralUS | ✓ |
southeastasia |
东南亚 | ✓ |
southindia |
南印度 | ✓ |
swedencentral |
瑞典Central | ✓ |
switzerlandnorth |
瑞士北部 | ✓ |
switzerlandwest |
瑞士西部 | |
uaenorth |
阿联酋北部 | ✓ |
uksouth |
UKSouth | ✓ |
ukwest |
UKWest | ✓ |
westcentralus |
WestCentralUS | ✓ |
westeurope |
西欧 | ✓ |
westindia |
WestIndia | |
westus |
WestUS | ✓ |
westus2 |
WestUS2 | ✓ |
westus3 |
WestUS3 | ✓ |
无服务器可用性
区域 | 位置 | 无服务器 SQL 仓库 | 笔记本和作业的无服务器计算 | Mosaic AI 模型训练 – 预测 | 无服务器 Lakeflow 声明性数据管道 | 私有连接 | Databricks 应用 | Lakeflow Connect 中的托管连接器 |
---|---|---|---|---|---|---|---|---|
australiacentral |
澳大利亚中部 | |||||||
australiacentral2 |
AustraliaCentral2 | |||||||
australiaeast |
澳大利亚东部 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
australiasoutheast |
澳大利亚东南部 | |||||||
brazilsouth |
巴西南部 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
canadacentral |
CanadaCentral | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
canadaeast |
加拿大东部 | |||||||
centralindia |
CentralIndia | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
centralus |
CentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
chinaeast2 |
ChinaEast2 | |||||||
chinaeast3 |
ChinaEast3 | |||||||
chinanorth2 |
ChinaNorth2 | |||||||
chinanorth3 |
ChinaNorth3 | |||||||
eastasia |
EastAsia | ✓ | ✓ | ✓ | ✓ | ✓ | ||
eastus |
EastUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
eastus2 |
EastUS2 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
francecentral |
FranceCentral | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
germanywestcentral |
德国西中部 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
japaneast |
JapanEast | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
japanwest |
日本西部 | |||||||
koreacentral |
KoreaCentral | ✓ | ✓ | ✓ | ✓ | ✓ | ||
mexicocentral |
墨西哥中心 | |||||||
northcentralus |
NorthCentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
northeurope |
NorthEurope | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
norwayeast |
挪威东部 | |||||||
qatarcentral |
卡塔尔中央 | |||||||
southafricanorth |
南非北 | |||||||
southcentralus |
SouthCentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
southeastasia |
东南亚 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
southindia |
南印度 | |||||||
swedencentral |
瑞典Central | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
switzerlandnorth |
瑞士北部 | ✓ | ✓ | ✓ | ✓ | ✓ | ||
switzerlandwest |
瑞士西部 | |||||||
uaenorth |
阿联酋北部 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
uksouth |
UKSouth | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
ukwest |
UKWest | |||||||
westcentralus |
WestCentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ||
westeurope |
西欧 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
westindia |
WestIndia | |||||||
westus |
WestUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
westus2 |
WestUS2 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
westus3 |
WestUS3 | ✓ | ✓ | ✓ | ✓ | ✓ |
AI 和机器学习
区域 | 位置 | 矢量搜索 | 预测性优化 | Mosaic AI 框架和评估 | 基础模型微调 | 针对批处理推理优化的 AI 函数 | 湖屋监视 |
---|---|---|---|---|---|---|---|
australiacentral |
澳大利亚中部 | ||||||
australiacentral2 |
AustraliaCentral2 | ||||||
australiaeast |
澳大利亚东部 | ✓ | ✓ | ✓ | ✓† | ✓ | |
australiasoutheast |
澳大利亚东南部 | ||||||
brazilsouth |
巴西南部 | ✓ | ✓ | ✓ | ✓† | ✓ | |
canadacentral |
CanadaCentral | ✓ | ✓ | ✓ | ✓† | ✓ | |
canadaeast |
加拿大东部 | ||||||
centralindia |
CentralIndia | ✓ | ✓ | ✓ | ✓† | ✓ | |
centralus |
CentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
chinaeast2 |
ChinaEast2 | ||||||
chinaeast3 |
ChinaEast3 | ||||||
chinanorth2 |
ChinaNorth2 | ||||||
chinanorth3 |
ChinaNorth3 | ||||||
eastasia |
EastAsia | ✓ | ✓ | ✓† | ✓ | ||
eastus |
EastUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
eastus2 |
EastUS2 | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
francecentral |
FranceCentral | ✓ | ✓ | ✓ | ✓ | ✓ | |
germanywestcentral |
德国西中部 | ✓ | ✓ | ✓ | ✓ | ✓ | |
japaneast |
JapanEast | ✓ | ✓ | ✓ | ✓† | ✓ | |
japanwest |
日本西部 | ||||||
koreacentral |
KoreaCentral | ✓ | ✓ | ✓† | ✓ | ||
mexicocentral |
墨西哥中心 | ||||||
northcentralus |
NorthCentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
northeurope |
NorthEurope | ✓ | ✓ | ✓ | ✓ | ✓ | |
norwayeast |
挪威东部 | ||||||
qatarcentral |
卡塔尔中央 | ||||||
southafricanorth |
南非北 | ||||||
southcentralus |
SouthCentralUS | ✓ | ✓ | ✓ | ✓ | ✓ | |
southeastasia |
东南亚 | ✓ | ✓ | ✓ | ✓† | ✓ | |
southindia |
南印度 | ||||||
swedencentral |
瑞典Central | ✓ | ✓ | ✓ | ✓ | ✓ | |
switzerlandnorth |
瑞士北部 | ✓ | ✓ | ✓ | ✓ | ||
switzerlandwest |
瑞士西部 | ||||||
uaenorth |
阿联酋北部 | ✓ | ✓ | ✓ | ✓† | ✓ | |
uksouth |
UKSouth | ✓ | ✓ | ✓ | ✓ | ✓† | ✓ |
ukwest |
UKWest | ✓ | ✓ | ||||
westcentralus |
WestCentralUS | ✓ | |||||
westeurope |
西欧 | ✓ | ✓ | ✓ | ✓ | ✓ | |
westindia |
WestIndia | ||||||
westus |
WestUS | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
westus2 |
WestUS2 | ✓ | ✓ | ✓ | ✓ | ||
westus3 |
WestUS3 | ✓ | ✓ | ✓ | ✓ |
仅 当启用了跨地域路由时,才支持†区域。 依赖于 GPU 计算可用性的基础模型 API 工作负载(例如使用 Meta Llama 系列中的模型的工作负载)需要跨地理位置路由。
模型服务区域可用性
下表汇总了马赛克 AI 模型服务功能的区域可用性。
重要
对于基础模型 API(按令牌付费):
- Meta Llama 4 Maverick、GTE 大型(En)和 Meta Llama 3.3 70B 指令模型以按令牌计费模式在欧盟和美国支持的地区提供。 请参阅 Databricks 上托管的基础模型 ,这些模型在可用的按令牌付费区域中受支持。
- 如果工作区位于“模型服务”区域,但不位于“美国”或“欧盟”区域中,则工作负荷将路由到“美国 Databricks Geo”。 有关如何启用跨区域数据处理,请参阅 “启用跨地理位置处理”。 请参阅 Databricks 指定服务,以了解各地区如何处理按令牌计价的工作负载。
对于基础模型 API 预配吞吐量,可以从 HuggingFace 下载 Meta Llama 3.1 和 3.2 模型,并使用预配的吞吐量终结点提供服务。 有关受支持的区域和模型变体,请参阅 Databricks 上托管的基础模型 。
区域 | 位置 | 核心模型服务功能 ** | 外部模型 | 基础模型 API (预配吞吐量) ** | 基础模型 API(按代币付费) |
---|---|---|---|---|---|
australiacentral |
澳大利亚中部 | ||||
australiacentral2 |
AustraliaCentral2 | ||||
australiaeast |
澳大利亚东部 | ✓ | ✓ | ✓ | ✓† |
australiasoutheast |
澳大利亚东南部 | ||||
brazilsouth |
巴西南部 | ✓ | ✓ | ✓ | ✓† |
canadacentral |
CanadaCentral | ✓ | ✓ | ✓ | ✓† |
canadaeast |
加拿大东部 | ||||
centralindia |
CentralIndia | ✓ | ✓ | ✓ | ✓† |
centralus |
CentralUS | ✓ | ✓ | ✓ | ✓ |
chinaeast2 |
ChinaEast2 | ||||
chinaeast3 |
ChinaEast3 | ||||
chinanorth2 |
ChinaNorth2 | ||||
chinanorth3 |
ChinaNorth3 | ||||
eastasia |
EastAsia | ✓ | ✓ | ✓† | ✓† |
eastus |
EastUS | ✓ | ✓ | ✓ | ✓ |
eastus2 |
EastUS2 | ✓ | ✓ | ✓ | ✓ |
francecentral |
FranceCentral | ✓ | ✓ | ✓ | ✓ |
germanywestcentral |
德国西中部 | ✓ | ✓ | ✓ | ✓ |
japaneast |
JapanEast | ✓ | ✓ | ✓† | ✓† |
japanwest |
日本西部 | ||||
koreacentral |
KoreaCentral | ✓ | ✓ | ✓† | ✓† |
mexicocentral |
墨西哥中心 | ||||
northcentralus |
NorthCentralUS | ✓ | ✓ | ✓ | ✓ |
northeurope |
NorthEurope | ✓ | ✓ | ✓ | ✓ |
norwayeast |
挪威东部 | ||||
qatarcentral |
卡塔尔中央 | ||||
southafricanorth |
南非北 | ||||
southcentralus |
SouthCentralUS | ✓ | ✓ | ✓ | ✓ |
southeastasia |
东南亚 | ✓ | ✓ | ✓† | ✓† |
southindia |
南印度 | ||||
swedencentral |
瑞典Central | ✓ | ✓ | ✓ | ✓ |
switzerlandnorth |
瑞士北部 | ✓ | ✓ | ✓ | ✓ |
switzerlandwest |
瑞士西部 | ||||
uaenorth |
阿联酋北部 | ✓ | ✓ | ✓† | ✓† |
uksouth |
UKSouth | ✓ | ✓ | ✓† | ✓† |
ukwest |
UKWest | ||||
westcentralus |
WestCentralUS | ||||
westeurope |
西欧 | ✓ | ✓ | ✓ | ✓ |
westindia |
WestIndia | ||||
westus |
WestUS | ✓ | ✓ | ✓ | ✓ |
westus2 |
WestUS2 | ✓(预览) | ✓(预览) | ✓ | ✓(预览) |
westus3 |
WestUS3 | ✓ | ✓ | ✓ | ✓ |
**包括 GPU 支持
仅 当启用了跨地域路由时,才支持†区域。 依赖于 GPU 计算可用性的基础模型 API 工作负载(例如使用 Meta Llama 系列中的模型的工作负载)需要跨地理位置路由。