Here at Ampere we've built an extensive infrastructure focused on Continuous Integration, Delivery, and Regression (called CIDR).

[Read More About CIDR](drawer:/cidr-overview)

[CIDR Demo](https://uawartifacts.blob.core.windows.net/upload-files/Ampere_CIDR_fc459cb2f4.mp4)

Our testing runs 24/7/365 in our regression infrastructure. 

[Basic Functional Test](drawer:/basic-functional-tests)

Results are categorized as either 'Verified' or 'Unverified'.

[Know More](drawer:/cidr-verified-unverified)

Test and build infrastructure can encounter complexities or unexpected speed bumps. Known incidents and their resolutions will be documented where applicable.

[DockerHub Introduces Image Pull Limits (2020.11)](drawer:/dockerhub-image-pull-limits)

Date,Platform and OS details of verified test results are displayed on hover bubble.

To give feedback on this portal please send us a message at 

[developer@amperecomputing.com](mailto:developer@amperecomputing.com)

Redis tests were performed on bare-metal single socket servers with equivalent memory, networking and storage configurations for each of the platforms shown. The processors tested here are: AMD EPYC 7763 "Milan"; Intel Xeon 8380 "Ice Lake"; Ampere Altra Q80-30; Ampere Altra Max M128-30.

This test was performed using the Memtier_benchmark (developed by Redis Labs) as a load generator for benchmarking Redis. Each test was configured to run with multiple threads, multiple clients per thread, and with pipelining enabled.

It is recommended to compile Redis server with GCC (GNU Compiler Collection) 10.2 or newer as newer compilers have made significant progress towards generating optimized code that can improve performance for Aarch64 applications.

CentOS 8.3 (kernel 4.18) with Redis-server 5.0.12 were used. For each of the tests, a similar number of clients was used to generate requests to Redis-server.

Since it is realistic to measure throughput under a specified Service Level Agreement (SLA), a 99th percentile latency (p.99) of 1 millisecond was used. This ensured that 99 percent of the requests had a response time of 1 ms in the worst case.

The test ran for 3 minutes with a 1:10 get:set ratio, which is common for in-memory caches. Sn appropriate number of clients and threads/client to load one instance of Redis was initially used, while ensuring the p.99 latency was at most 1 ms. Pipelining is a feature whereby Redis can process new requests even if the client has not already read older responses. This feature can dramatically reduce response times and 45 concurrent pipelined requests were used.

Next, the number of Redis instances was successively increased till one or more instances violated the p.99 latency SLA. The aggregate throughput of all instances was used as the primary performance metric. The test was run three times and minimal run-to-run variation was observed

Redis Test Notes

Basic Functional Tests

While performance testing is a core part of our DNA, the Basic Functional Test is the first thing we look at. This functional test will brings up a system, install the complete software package, and measure a collection of Hello World type tests.

Each test is hand written to the specific piece of software under test and ensures that everything installs and runs as expected. 

The Basic Functional Test is not complex, but it is essential.

TEST DOCUMENTATION

CIDR Overview

Here at Ampere we believe in full automation and continuous best practices for all of our development activities. To meet our goals of a fully automated software environment we built a system called CIDR: Continuous Integration, Delivery, and Regression. 

CIDR is a framework composed of industry-best tools sitting on top of a robust database and analysis layer. This is all built using Python, Django, Postgres, serverless cloud environments, and more. We use CIDR to ensure that all of our systems and software meet the rigorous quality needs of the modern data center. 

Our CIDR data analysis tools are purpose-built to handle both functional and performance tests, giving us a complete picture of how our software performs on real hardware.

Results and data that you see in the CI/CD section of this portal are all collected using the CIDR infrastructure with Ampere bare-metal at its core. 

Find more details in our [All About CIDR](https://solutions.amperecomputing.com/posts/all-about-cidr) article.

CONTINUOUS INTEGRATION, DELIVERY, REGRESSION

Verified vs. Unverified

Results are categorized as either 'Verified' or 'Unverified'. Verified: Everything ran according to plan and results came back positive Unverified: We couldn't collect a result due to an issue within the test infrastructure. When we root-cause an Unverified result the write-up will appear in the Test Notes section of this page. An unverified test result does not imply an issue with the software under test - it means only that Ampere was unable to confirm one or more steps in our verification process.

Ampere Computing Solutions

DockerHub Image Pull Limits

On November 22nd, 2020 we noticed random failures across our DockerHub infrastructure. This was caused by a failure to authenticate with DockerHub following the rollout of rate limiting ([see here](https://docs.docker.com/docker-hub/download-rate-limit/)).

This presented with an error message similar to the following:

> *You have reached your pull rate limit. You may increase the limit by authenticating and upgrading...*

We rolled out a fix to authenticate with our DockerHub organization. 

This affected results collected between November 20th, 2020 and November 25th, 2020.

ROOT CAUSE ANALYSIS

Ampere Altra and Ampere Altra Max. These systems are flexible enough to meet the needs of any cloud deployment and come packed with Ampere's 80-core Altra or 128-core Altra Max processors

Ampere Altra Systems

Ampere Altra

Microsoft offers a comprehensive line of Azure Virtual Machines that can run a diverse and broad set of Linux workloads such as web servers, open-source databases, in-memory applications, big data analytics, gaming, media, and more.

Microsoft Azure

Azure

Equinix Metal, an on-demand digital infrastructure platform, has created Gen3 configs with Ampere Altra for common workloads which are available in minutes on bare metal

Equinix Metal

Equinix

Whether your business is early in its journey or well on its way to digital transformation, Google Cloud can help you solve your toughest challenges.

Google Cloud

The new HPE ProLiant RL300 Gen11 server is the first in a series of HPE ProLiant RL Gen11 servers that deliver next-generation compute performance with higher power efficiency using Ampere® Altra® and Ampere ® Altra® Max cloud-native processors.

Hewlett Packard Enterprise

Ampere Altra and Oracle Cloud combine predictable performance, near-linear scaling, and secure architecture with the best price-performance in the market in the following shapes:  
- Flexible Virtual Machines
- Bare Metal

OCI Ampere A1

Oracle

Tencent Cloud's new SR1 instance is now publicly available - offering scalable VM Instances (up to 70 cores) allows customers to match an instance to their workload needs

Tencent Cloud

Hi, this website uses essential cookies to ensure its proper operation, analytical cookies to improve its functionality, and tracking cookies to understand how you interact with it. The non-essential cookies will be set only after consent.

Name

These cookies allow us and our third-party service providers to recognize and count the number of visitors and to see how visitors move around our website when they are using it. This helps us improve how our website works by, for example, ensuring that users can easily find what they need on our website. These cookies generate aggregate statistics that are not associated with an individualized profile.

Performance and Analytics cookies

Our website requires the use of these cookies to properly operate or provide necessary functions relating to the services you request. They include, for example, cookies enabling us to track your preferences and allow third party embed functionality.

Strictly necessary cookies

With up to 192 cores,  AmpereOne® is the next generation of Cloud Native Processors for sustainable compute.

AmpereOne® Product Brief

Processors

Partners

Product Security

Where to Buy

Where to Try

Products

We have created a new class of GPU-free AI inferencing that delivers the best price/performance for many AI models.

Ampere AI Solutions

Solutions Center

Artificial Intelligence

ARM Native Apps

Data Analytics

Edge and Embedded

Media Services

Storage

Web Services

Solutions

How FreeBSD Employs Ampere Arm64 Servers in the Data Center

Ampere in the Wild:

Developer Center

Design

Build

Deploy

Power Your AI

Working with Open Source

Ampere Ready Software

Community Forum

Subscribe to Newsletter

Developers

The Data Center Dilemma: Balancing Growth, Efficiency & Environmental Impact

Blog

Events

Webinars

Case Studies

Blogs

Ampere YouTube

Briefs

Tutorials

Tuning Guides

Resources

Sandy Dash Engineer on the Silicon Validation team

Meet the Team at Ampere:

About Us

Leadership Team

Newsroom

Careers

Contact Us

Contact Sales

Company

Elasticsearch is a search and analytics engine. It is a search engine provided with distributed, multitenant-capability with an HTTP web interface. Elasticsearch is developed in Java. It centrally stores your data for faster search, finetuned relevancy, and scalable analytics. Elasticsearch allows you to perform and combine many types of searches. These searches include structured, unstructured, geo or metric searches. Elasticsearch lets you explore various trends and patterns in your data. It also provides rapid results, powerful designs, and indexing. Elasticsearch is followed by Apache Solr, based on Lucene.

Data shown:
- **Performance:** Throughput measured in total Requests per second
- **Power Efficiency:** More Throughput delivered per Watt 

Overview

Elasticsearch

System configurations, components, software versions, and testing environments that differ from those used in Ampere’s tests may result in different measurements than those obtained by Ampere.  The system configurations and components used in our testing are detailed [here](drawer:/redis_drawer)

Elasticsearch recommended systems, testing & regression with Ampere Cloud Native Processors, and developer resources for instances on OCI, Azure, HPE, and more.

Elasticsearch is a search and analytics engine. It is a search engine provided with distributed, multitenant-capability with an HTTP web interface. Elasticsearch is developed in Java. It centrally stores your data for faster search, finetuned relevancy, and scalable analytics. Elasticsearch allows you to perform and combine many types of searches. These searches include structured, unstructured, geo or metric searches. Elasticsearch lets you explore various trends and patterns in your data. It also provides rapid results, powerful designs, and indexing. Elasticsearch is followed by Apache Solr, based on Lucene.

Databases

Visit the Official Home Page

See the Image Sources

Visit the Official Docker Hub Page

Ampere Computing's platforms are uniquely designed to meet the needs of the modern cloud native workload. DockerHub hosts a number of official images for a wide range of software that can be pulled and used anywhere docker is supported. These are a set of images hand selected by a dedicated team at Docker, Inc. 

More information can be found in the official documentation on [Official Images on Docker Hub](https://docs.docker.com/docker-hub/official_images/).

Recommended Systems

Recommended Systems for Elasticsearch

Testing & Regression

Regression Testing Results

Resources

Resources for Elasticsearch