How to use hyper-convergence to improve the efficiency of manufacturing development and testing [attached user cases]

Entering the digital age, more and more manufacturing companies are accelerating application innovation through independent research and development. Since R&D efficiency largely depends on the performance of IT infrastructure such as servers and storage that support development and testing, enterprises need a simple, flexible, and high-performance IT infrastructure to provide resource sharing and rapid delivery support for R&D teams.

In response to this demand, SmartX provides manufacturing users with a high-performance, easy-to-expand hyper-converged infrastructure, which can not only build a simple and flexible shared resource pool for development and compilation, but also support the smooth operation of R&D virtual desktops. Below we will use two examples of user practices to introduce in detail how SmartX hyper-convergence can improve the efficiency of enterprise development and testing.

Case number one

A well-known smartphone manufacturer: Comparing with VMware virtualization + virtual desktop solution, SmartX hyper-converged compilation task execution efficiency increased by more than 50%

A well-known domestic smartphone manufacturer originally used physical machines to build a development and testing environment, and the utilization rate of hardware resources was low. At the same time, due to limited equipment, the development and testing team has been using the mechanism of "R&D personnel use personal computers for temporary compilation, each R&D team queues up to apply for procurement, and uses the server for formal compilation" before building a development and testing platform that can realize resource sharing. , seriously affecting R&D efficiency and business launch speed.

In order to solve this problem, the user originally planned to use the VMware virtualization + Citrix virtual desktop solution to upgrade the architecture. However, the operation and maintenance of this solution is complex, the investment cost is high, and it cannot fully meet the needs of users. Therefore, users put their hopes on the hyper-converged solution with a simpler architecture and higher cost performance, and conducted performance tests on the virtual desktop solution and the SmartX hyper-converged solution. The results show that under the same conditions, the efficiency of SmartX hyper-converged compilation tasks is improved by more than 50% compared with the original plan, which greatly exceeds user expectations.

In the end, the user decided to use the SMTX Halo hyper-converged all-in-one machine (based on SmartX's native virtualized ELF) to deploy 10 nodes, and build a development and testing platform with the Cloud-init function in the Ubuntu system. This solution not only solves the problem that users are difficult to share resources, improves R&D efficiency, but also saves the authorization cost of VMware virtualization and virtual desktop by choosing SmartX native virtualization ELF, and further improves the agility of R&D.

Case 2

CASCO: Using SmartX hyper-convergence to support the entire VDI system, reducing the running time of R&D programs by an average of 60%

As an all-round rail transit control system integrator, CASCO Signaling Co., Ltd. invests no less than 10% of its annual sales in research and development, and continues to innovate in the field of rail transit control systems. Users originally used IBM blade servers with IBM v7000 centralized storage to support 900+ VDI desktops, but as the number of VDI users increased rapidly at a scale of 100+ per year, the storage capacity gradually became saturated, and the performance gradually reached the upper limit of the storage controller . Problems such as slow compilation speed and unresponsive desktop have hindered the smooth development and testing. In addition, customer blade servers are also approaching their end of life, and business risk is increasing due to blade servers sharing hardware components. Therefore, users expect new architectures to support development and test scenarios.

After contacting with SmartX engineers, CASCO conducted three POC tests on the ability of SmartX hyper-convergence to support virtual desktop scenarios, respectively based on the old version of Citrix Xenserver + XenDesktop, the new version of Citrix Xenserver + XenDesktop and VMware virtualization + XenDesktop to verify the SmartX super Converged performance, compatibility and migration solutions. The test results show that the SmartX hyper-converged architecture with VMware virtualization performs best, and the R&D environment is higher than the original architecture in terms of performance and response speed:

  • The average running time of self-developed programs can be reduced by  60% .

  • The average opening time of development documents can be reduced by  40% .

  • The average VDI user login time can be reduced by  15% .

In view of the high performance and flexible expansion capabilities of SmartX hyper-convergence, the user finally chose to use SmartX hyper-convergence (based on VMware virtualization) to carry the entire R&D VDI application system (Citirix XenDesktop) and a set of SAP business systems for data center clusters, and gradually migrated 250 + Taiwan develops VDI desktop. Since its deployment two years ago, the entire architecture has been operating stably, greatly improving research and development efficiency, and supporting the expansion of resources as VDI users grow. In the future, the user plans to migrate all VDI desktops to the SmartX hyper-converged platform, and use the hyper-converged architecture as the underlying platform for future informatization construction to carry VDI, core production services, and build cloud laboratories.

 

For case details, please read: CASCO: Hyper-convergence speeds up the development of rail transit signaling systems.

In addition to development and testing scenarios, SmartX hyper-convergence can also fully support the cloud transformation of manufacturing infrastructure in scenarios such as core business, multi-factory management, and disaster recovery backup. To learn more about the plans and practices, please read:

How to build a simple, intelligent and reliable branch factory IT infrastructure?

How effective is the SmartX hyper-converged disaster recovery solution? Let’s look at 3 examples of manufacturing user practices

You can also scan the QR code below to download the e-book "IT Infrastructure Transformation Scheme and Practice in the Manufacturing Industry" to gain an in-depth understanding of hyper-convergence in the overall IT infrastructure transformation, core business and disaster recovery, multi-factory construction and management, and development of the manufacturing industry. Solutions and success stories for multiple scenarios such as testing and virtual desktops.

 

Understand the technology and features behind the solution

SmartX hyper-convergence's excellent support for development and test scenarios comes from its fully distributed architecture and self-developed distributed block storage: SmartX hyper-convergence is based on "SMTX OS hyper-convergence software + commercial server + standard Ethernet switch" , replacing the traditional architecture based on traditional virtualization and centralized storage, reducing storage pressure and expansion difficulty; end-to-end performance optimization (such as I/O localization, Boost mode, hot and cold data layering mechanism, using RDMA protocol) also This ensures that SmartX can run efficiently and smoothly even with large amounts of data processing and user login.

 

To learn more about how SmartX hyperconvergence optimizes storage performance, read:

Analysis of VMware and SmartX hyper-converged I/O path differences and their impact

Analysis and Performance Comparison of VMware and SmartX Distributed Storage Cache Mechanism

The independent research and development journey of distributed block storage ZBS|RDMA of data synchronization protocol

The independent research and development journey of distributed block storage ZBS|NVMe-oF of access protocol

How SPDK Vhost-user helps hyper-converged architecture improve I/O storage performance

At the same time, combined with SmartX hyper-converged elastic expansion, high availability, simple operation and maintenance and other features, R&D and business delivery efficiency can be further improved:

  • Good elastic expansion capability : starting with 3 nodes, the capacity of the resource pool can be expanded on demand as the business develops , and the performance also increases linearly with the expansion of resources, which meets the needs of enterprise sensitive business development.
  • Multiple desktop virtualization support : SmartX hyper-converged solutions can be combined with Citrix XenServer and VMware vSphere to support Citrix Virtual Apps and Desktops and VMware Horizon View, meeting various customer needs.
  • Comprehensive availability assurance : Ensure infrastructure availability through multi-level disaster recovery solutions such as cluster high availability, backup and recovery, asynchronous replication, and active-active clusters , so as to avoid business interruptions, data loss, and other situations that affect the progress of development projects.
  • Simple operation and maintenance experience : SmartX's hyper-converged and streamlined architecture and the decoupling of software and hardware greatly simplify daily operation and maintenance. At the same time, with a variety of intelligent functions and the native management platform  CloudTower , developers can focus on innovation.

Guess you like

Origin blog.csdn.net/weixin_43696211/article/details/130524638