首页> 外文会议>Conference on Software and Cyberinfrastructure for Astronomy >Enabling faster ALMA software delivery by using containers
【24h】

Enabling faster ALMA software delivery by using containers

机译:通过使用容器实现更快的ALMA软件交付

获取原文

摘要

ALMA hasbeen operating since 2013 and it keeps on adding an ever-growing new set of capabilities. Every new feature implies amongother things a new software release thathasto be implemented, tested and deployed around the world. In this paper we present the new deployment process that allowed ALMA to deliver faster releases to reliable testing and production environments. This was achieved through the use of container-based services, both for applications and data. This implied that tasks that in thepast were done manually, are nowfully automated, in orderto avoid human errors and maintain consistency between what was tested and what is finally installed in production. All this, under a unique and complex operation environment that includes the two main operation facilities in Chile at ALMA Operation Site Facilities (OSF) and Santiago Central Office (SCO), and the different executive headquarters located acrossthe ALMA global network. We also explain how we managed to address the issue of ever-growing observational data, which made it difficult to replicate the production environment data into ourtesting infrastructure. Our solution consisted in using a container-based database that allowed us to create a full copy of the production database in a very short time. All those changes enabled JAO to improve its software testing process allowing a monthly release cycle.
机译:自2013年以来,Alma从2013年开始运营,它一直在增加一个不断增长的新功能。每一个新功能都意味着在世界各地实施,测试和部署的新软件发布的新功能。在本文中,我们介绍了新的部署过程,允许Alma提供更快的释放到可靠的测试和生产环境。这是通过使用基于集装箱的服务来实现的,包括应用程序和数据。这意味着在手动完成的任务,现在是自动化的,以避免人类错误并在测试的内容之间保持一致性以及最终安装在生产中的一致性。所有这一切,在一个独特而复杂的运营环境下,包括智利的两个主要运营设施在阿尔马运营网站设施(OSF)和Santiago Central Office(SCO),以及位于Actossthe Alma Global Network的不同行政总部。我们还解释了我们如何设法解决不断增长的观察数据问题,这使得难以将生产环境数据复制到我们的基础架构中。我们的解决方案在使用基于容器的数据库中,使我们能够在很短的时间内创建生产数据库的完整副本。所有这些更改都使能JAO改进其软件测试过程,允许每月发布周期。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号