All Products
Search
Document Center

Elastic High Performance Computing:Release notes

Last Updated:Jan 04, 2024

This topic describes the release notes for Elastic High Performance Computing (E-HPC) and provides links to the relevant references.

2023

September 2023

Feature

Type

Description

Release date

Reference

Job submission by using a serverless cluster

New feature

The compute nodes of a serverless cluster consist of elastic container instances. You only need to provide packaged container images to use elastic container instances to run jobs.

2023-09-08

August 2023

Feature

Type

Description

Release date

Reference

Supported schedulers

Optimized feature

The schedulers supported by the cluster are updated.

2023-08-29

Cluster overview

June 2023

Feature

Type

Description

Release date

Reference

Job submission by using CLIs

Optimized feature

The sample commands are optimized.

2023-06-21

Use the CLI to submit a job

Cluster creation by using a template

Optimized feature

The parameters that you need to configure when you create a cluster by using a template are updated.

2023-06-14

Create a cluster by using a template

May 2023

Feature

Type

Description

Release date

Reference

Molecular structure calculation by using Schrodinger

New feature

Best practice on how to use Schrodinger in an E-HPC cluster to calculate molecular structure is added.

2023-05-19

Use Schrodinger to calculate molecular structure

Disabling HT for compute nodes

New feature

Each compute node in an E-HPC cluster is an Elastic Compute Service (ECS) instance. By default, Hyper-Threading (HT) is enabled for each ECS instance. In some high-performance computing (HPC) scenarios, you can disable HT to improve the performance of instances.

2023-05-08

Disable HT for compute nodes

March 2023

Feature

Type

Description

Release date

Reference

Ubuntu support for E-HPC clusters

New feature

Public image for Ubuntu 20.04 is released. E-HPC clusters of the Ubuntu operating system can be created.

2023-03-15

Release notes in 2022

December 2022

Feature

Type

Description

Release date

Reference

Automatic RAM role binding to added nodes

New feature

If you authorized RAM roles in the Authorized Instance Configuration section during cluster creation, the RAM roles are automatically bound to the nodes that are added during cluster scale out.

2022-12-07

N/A

October 2022

Feature

Type

Description

Release date

Reference

E-HPC clusters equipped with Yitian chips

New feature

E-HPC clusters equipped with Yitian chips can be created. A complete HPC software stack is provided, including the HPC scheduler, MPI library, and HPC software.

The g8m instance family is in invitational preview. You can go to the g8m Instance Free Trial Application Form page to apply for a free trial use.

2022-10-26

Create a cluster by using the wizard

September 2022

Feature

Type

Description

Release date

Reference

Operation logs

Optimized feature

The query of operation logs is optimized:

  • Operation logs can be queried by time range, operation type, result, or request ID.

  • Operation logs of the last 30 days can be queried. The time range to query can be up to 7 days.

2022-09-01

View E-HPC operations logs

Global search in the E-HPC console

New feature

E-HPC clusters and node resources can be queried in the top search bar of the E-HPC console. In the search result list, you can click a resource to open the corresponding page to view the resource details.

2022-09-01

N/A

August 2022

Feature

Type

Description

Release date

Reference

Reports of E-HPC Slurm clusters

New feature

The core hours of E-HPC Slurm clusters can be queried by user or queue.

2022-08-31

N/A

June 2022

Feature

Type

Description

Release date

Reference

E-HPC client

New feature

An E-HPC client can be used to submit jobs, query jobs, and manage job files. This way, you can manage jobs with enhanced efficiency and user experience.

2022-6-16

Overview

Job monitoring

New feature

The number of running, submitted, queuing, and completed jobs can be viewed in the E-HPC console to monitor the jobs.

2022-06-07

Monitor jobs

May 2022

Feature

Type

Description

Release date

Reference

E-HPC cluster creation

Optimized feature

Software can no longer be installed when you use the wizard of the E-HPC console to create an E-HPC cluster. This helps you create E-HPC clusters with higher efficiency. After the cluster is created, you can install software in the Software Manage panel based on your needs.

2022-05-16

Create a cluster by using the wizard

April 2022

Feature

Type

Description

Release date

Reference

Storage resource monitoring

New feature

The cluster storage performance and storage usage of a time period can be viewed to monitor the status of storage resources.

2022-04-27

Monitor storage resources

Auto scaling

Optimized feature

The maximum and minimum number of compute nodes that you want to add in each round of scale-out can be configured.

2022-04-01

Configure an auto scaling policy

March 2022

Feature

Type

Description

Release date

Reference

User management

Optimized feature

The console permissions of RAM users can be modified. This feature is used to specify the modules that are visible to RAM users after they log on to the E-HPC console.

2022-03-15

Change the console permissions of a RAM user

Releases before 2022

Release notes in 2021

Feature

Type

Description

Release date

Reference

Hybrid cloud cluster

Optimized feature

The proxy mode is supported for hybrid cloud clusters. The proxy nodes on the cloud can be used to manage cloud cluster resources for on-premises clusters.

2021-12-07

Create a hybrid cloud cluster

Job submission

Optimized feature

The Submit Job tab of the Job module is updated. Workbench can be used to log on to E-HPC clusters and edit job files.

2021-11-25

Use the E-HPC console to submit a job

Hybrid cloud cluster scale-out

Optimized feature

Hybrid cloud clusters can be scaled out within a zone or across multiple zones based on your needs.

2021-10-26

Scale out a hybrid cloud cluster

User management

New feature

A RAM user can be bound to an E-HPC cluster to facilitate cluster management.

2021-09-29

Scheduler management

New feature

Scheduler settings, such as queue resource limits and job scheduling cycles, can be configured to improve resource usage.

2021-09-17

Configure scheduler settings

Homepage customization

New feature

The Cluster, Job, or Nodes page can be configured as the homepage of the E-HPC console.

2021-08-30

N/A

New region

New feature

E-HPC is available in the China (Heyuan) region.

2021-05-13

Limits

New region

New feature

E-HPC is available in the China (Ulanqab) region.

2021-03-12

Limits

Release notes in 2020

Feature

Type

Description

Release date

Reference

E-HPC cluster creation

Optimized feature

The file system type can be selected when you create an E-HPC cluster.

2020-11-23

Create a cluster by using the wizard

Hybrid cloud cluster

New feature

Open Grid Scheduler is supported for hybrid cloud clusters.

2020-06-02

Create a hybrid cloud cluster

Image change

New feature

The default image of a cluster can be changed in the E-HPC console.

2020-05-07

View and modify cluster information

Template-based E-HPC cluster creation

New feature

E-HPC clusters can be created based on a template. Template files can be imported and exported.

2020-04-24

Create a cluster by using a template

2018

Feature

Type

Description

Release date

Reference

Auto scaling

Optimized feature

Auto scaling can be configured by queue for an E-HPC cluster.

2018-09-27

Configure an auto scaling policy

GPU performance monitoring

New feature

The GPU performance of an E-HPC cluster can be monitored.

2018-09-27

View cluster performance

E-HPC cluster scale-out

Optimized feature

The compute nodes of an instance type that is different from the existing instance type can be added to an E-HPC cluster when you scale out the cluster.

2018-07-12

Scale out a cluster

Job submission

New feature

Jobs can be submitted in the E-HPC console.

2018-06-29

Use the E-HPC console to submit a job

Auto scaling

Optimized feature

Auto scaling is supported by the E-HPC clusters that use the LSF scheduler.

2018-06-05

Configure an auto scaling policy

Auto scaling

Optimized feature

Preemptible instances are supported by auto scaling policies.

2018-05-22

Configure an auto scaling policy

Billing method

New feature

Subscription E-HPC clusters can be created and scaled out.

2018-04-26

Billing overview

E-HPC cluster status

New feature

The Stopped state and Deploying state are added for E-HPC clusters.

2018-04-18

View and modify cluster information

New region

New feature

E-HPC is available in the China (Qingdao) region.

2018-01-25

Limits

E-HPC

New feature

E-HPC is released.

2018-01-11

What is E-HPC?