SQL Server 2005 +生产中一台计算机上的 JavaEE 应用程序
我们有一个服务器项目，每天大约有 3000 个来自客户端的请求。 30-40个在线用户。每年数据增长约 30-40Gb。 并且系统应该7×24小时在线。
我们在哪里可以更好地定位SQL服务器？ 在一台带有 JavaEE 容器的计算机上还是在独立计算机上？
这种方式有什么优点和缺点？
We have project with a server witch have about 3000 requests at day from client.
30-40 on-line users. And smth about 30-40Gb data increase per year.
And system should be on-line 24x7.
Where can we locate the SQL server better?
On one computer with JavaEE container or on standalone computer?
What pros and cons on this ways?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（3）
当然，正如 Cade Roux 所解释的那样，最好将它们放在单独的服务器上（如果它们通过 LAN（而不是 Internet）连接）。
但是，如果您的预算不足以购买两台计算机，或者您已经拥有一台功能强大的计算机，则可以使用虚拟化。
创建 2 台虚拟机，一台用于 Java EE 容器，一台用于数据库服务器。这通过一台物理机器提供了 Cade Roux 所解释的所有好处。
Of course, as Cade Roux explained, it's always better to put them on separate servers (in case they connected via LAN (and not Internet)).
If, however, your budget isn't enough for two machines, or you already have a powerfull machine, you can use virtualization.
Make 2 virtual machines, one for the Java EE container, one for the database server. This gives all the benefits explained by Cade Roux, with a single physical machine.
上述所有海报建议都是正确的。
然而，每天 3000 个请求听起来并不算多。
当然，如果您可以为每个服务器拥有一个单独的服务器，那么出于上述原因，那就更好了。如果没有，并且您的计算机足以运行这两者，您可以通过运行虚拟机来实现单独计算机的一些优点，但如果您的硬件崩溃或操作系统崩溃，虚拟机将不会为您购买任何东西。
All of the above posters suggestions are correct.
However, 3000 requests a day does not sound like much.
Of course, if you can have a separate server for each, it is much better for the reasons specified above. If not, and your machine is adequate to run both, you can achieve some of the advantages of separate machines by running a VM, but if your hardware crashes or OS crashes, the VM won't buy you anything.
通常，应用程序服务器组件不安装在数据库服务器上。此外，您的 24x7 需求可能要求您不要将应用程序层和数据库服务放在同一台计算机上，以便您可以分离数据库层，以便能够在数据库层实现某种集群或其他高可用性解决方案以及应用程序服务层的类似负载平衡或其他内容。
至于体积，从表面上看，一台机器处理起来似乎并不算多，但你确实没有给出足够的细节。
Typically, application server components are not installed on database servers. In addition, your 24x7 requirements probably dictate that you not put both the application layer and the database services on the same machine so that you can separate the database layer to be able to implement some kind of clustering or other high availability solution at the database layer and a similar load balancing or whatever at the application services layer.
As far as volume, on the face of it, it doesn't seem that it would be too much for one machine to handle, but you really haven't given enough detail.