Instance Type
|
| | Equipped with massive storage resources and characterized by high throughput, suitable for throughput-intensive applications such as Hadoop distributed computing, massive log processing, distributed file systems, and large-scale data warehouses |
Big Data Type Instance Family
Big Data Type DA4m
The DA4m instance, which is the latest generation of big data instances, is equipped with high throughput and massive storage resources. It can support up to 576 TB of SATA HDD local storage and is suitable for throughput-intensive businesses such as Hadoop distributed computing and parallel data processing.
Use Cases
Hadoop MapReduce/HDFS/Hive/HBase and other distributed computing technologies.
Designed for business scenarios such as Elasticsearch, log processing, and Large data warehouses.
Business scenarios involving massive data storage and computation for industry clients in sectors such as the Internet and finance, which have demands for big data computation and storage analysis..
Instance Characteristics
Based on Tencent Cloud's self-developed Star Lake servers, we provide reliable, secure, and stable high-performance services.
Using AMD EPYC™ Milan processors, with a base frequency of 2.55 GHz and a boost frequency of 3.5 GHz.
Instances can be equipped with up to 48 12 TB local disks, providing up to 576 TB of HDD-based local storage.
The processor-to-memory ratio is 1:4, designed for Big Data scenarios.
Supports up to 100 G private network bandwidth and 28,000,000 PPS. The network performance of an instance corresponds to its specifications: the higher the specification, the stronger the network forwarding performance, and the higher the private network bandwidth limit.
Supports disabling or enabling Hyper-Threading.
Instance Requirements
Big Data Type DA4m instances can be used as monthly subscription instances or pay-as-you-go instances.
DA4m instances can only be started within a Virtual Private Cloud.
DA4m instances do not support adjusting configuration.
The instance supports up to 100 G of network bandwidth, depending on the kernel version and running environment of the instance. When the PPS exceeds 10,000,000 and the bandwidth is greater than 50 Gbps, the kernel protocol stack can significantly impact network performance. In such cases, the bandwidth value from netperf testing may not meet expectations. You can use the DPDK method to shield the differences in the CVM's kernel protocol stack and obtain the true network performance of the instance. For the testing method, see High-throughput network performance testing. DA4m instances support purchasing configurations. See the instance specifications below. Ensure the DA4m instance size you select meets the minimum CPU and memory requirements of your operating system and application.
|
DA4m.2XLARGE32 | 8 | 32 | 1,700,000 | 1,000,000 | 8 | 6 | Equipped with 3 12 TB SATA HDD Local Disks |
DA4m.4XLARGE64 | 16 | 64 | 3,500,000 | 2,000,000 | 16 | 13 | Equipped with 6 12 TB SATA HDD Local Disks |
DA4m.8XLARGE128 | 32 | 128 | 7,000,000 | 4,000,000 | 32 | 25 | Equipped with 12 12 TB SATA HDD Local Disks |
DA4m.16XLARGE256 | 64 | 256 | 14,000,000 | 8,000,000 | 48 | 50 | Equipped with 24 12 TB SATA HDD Local Disks |
DA4m.32XLARGE512 | 128 | 512 | 28,000,000 | 16,000,000 | 48 | 100 | Equipped with 48 12 TB SATA HDD Local Disks |
Was this page helpful?