span8
span4
span8
span4
In 2017 we introduced theFME云实例的下一代coming withNGINX反向代理配置, more available regions, new instance types and the temporary disks.
To be able to fully use the power of the temporary disks we need to look into and understand 3 different concepts or mechanisms:
After understanding the implications and effects of these 3 aspects we will be able to exploit the temporary disk on FME Cloud to give our workflows a performance boost.
On the primary disk, we store the FME Server installation and the FME Server Shared resources with one exception, the Temp folder.The Temp folder in the FME Server Resources is a resource connection to the Temporary Disk of your FME Cloud instance.
As the name already implies the Temp folder should only be used for data that does not need to persist on your FME Cloud instance.Data that does not need to persist is usually uploaded user data that is processed in one of your workspaces, data that you might download from a cloud resource like AWS S3 to process on FME Cloud and data FME needs to write to disk during a transformation (we'll look into more details about this soon).There are a couple of reasons why this data should go to the temporary disk instead of the primary disk:
1。稳定性: When the primary disk is full, the Web Application server might shut down and fail to start up correctly even after an instance reboot.Often the only way to recover is a rollback to a previous backup.This risk can be avoided if we do not store data that is hard to control and estimate in size like user data or temporary FME output, on the primary disk.
2。灵活性: The temporary disk is purged after every reboot of the FME Cloud instance.This allows us to adjust the size of the temporary disk in a much more flexible way than the primary disk.The primary disk can only be increased in size, while the temporary disk can be increased and decreased in size to match your requirements at all times.
3。可用性: Because of the fact that the temporary disk is purged every time the FME Cloud instance has rebooted, the overhead to make sure enough disk space is available is significantly lower as you can purge the whole disk easily for example with a scheduled restart on the weekend.
4。承受能力: The temporary disk is not included in the FME Cloud backup and therefore you can dramatically lower your backup costs by storing the data you won't necessarily need in the future on the temporary disk.
很多与FME需要变换将数据写入到磁盘以备将来处理。其中最常见的原因就是FME Server尝试优化内存内存不足的情况。有些人可能会记得,以前在这样的日志文件的消息几年意味着您的工作基本上已经失败,即使它仍然在努力:
ResourceManager的:优化内存使用。请耐心等待...
但我们自那以后很长的路要走。我们的开发人员优化的内存利用率颇有几分也快了SSD硬盘的承受能力增加,使这种情况下没有绝望了,因为它曾经是。
对于内存密集型的转换通常的嫌疑人与合并或加铺后缀(如FeatureMerger或PointOnAreaOverlayer),点云相关的变压器和一般变压器,互感器是需要“HOLD功能”两种排序或分组的目的。
FME云运行在AWS EC2实例,并利用AWS通用SSD(GP2),卷。了解这些卷的两个重要的细节是I / O学分和突发事件的性能。一般情况下,我们可以说越大磁盘每秒更高的最小输入/输出操作(IOPS)。
此外,在基准性能,AWS的GP2容量大于1000 GB时,也能够在性能爆3000个IOPS基于信用体系上,那也是基于磁盘的大小:越大盘的时间越长 maximum burst performance duration.对于所有关于自动气象站GP2卷的细节表现,一定要看看这个AWS的博客文章:
在简化的方式,这意味着更大的我们临时磁盘的速度越快,我们可以读出和写入。而现在,我们知道,FME是利用临时磁盘中的转换,而我们更愿意存储输入数据从用户或云服务未来也临时磁盘上写我们的临时数据,很容易理解的能力来调整 the disk at our disposal is not only a tool to increase storage but also a great way to improve performance.
同时请参阅我们的文档和教程,了解如何调整你的临时磁盘的详细信息:
FME Cloud: How to Read the FME Cloud Metrics
How to keep the OS of your FME Cloud instance up to date
FME Cloud: Set up the 'Unresponsive server' alert
Moving from FME Server to FME Cloud
Getting Started with FME Cloud: Sign Up
Tutorial: Getting Started with FME Cloud
Getting Started with FME Cloud: Launch Instance
Working with Amazon S3 and FME
© 2019 Safe Software Inc |Legal