DTS、SSIS、Informatica、PL/SQL 脚本
过去，我曾使用Informatica进行一些ETL（提取转换加载），但发现它相当慢，通常用一些PL/SQL脚本替换它（当时使用Oracle）。
（根据答案中的反馈修改问题）
我认为 DTS 是 Microsoft 在 SSIS 之前的 ETL 工具。
- 使用 DTS 将现有应用程序转换为 SSIS 会很困难吗？
- 鉴于 SSIS 是一个 Microsoft 工具，并且与 SQL Server（实际上是 SQL Server 的一部分）紧密集成，使用它有什么缺点吗？我没有看到任何效率问题，因为我认为在 SSIS 中您可以做任何关于 ETL 的事情，而无需 SSIS。
In the past, I have used Informatica for some ETL (Extraction Transformation Loading) but found it rather slow and usually replaced it with some PL/SQL scripts (was using Oracle at the time).
(questions revised based on feedback in answers)
I gather that DTS was Microsoft's ETL tool prior to SSIS.
- Would it be difficult to convert an existing application using DTS to SSIS?
- Given that SSIS is a Microsoft tool and tightly integrated with SQL Server (virtually a part of it) are there any drawbacks to using it? I don't see any efficiency issues, since I imagine that you can do anything in SSIS that you could without it with regard to ETL.
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（4）
我相信 SSIS 是微软今天的 ETL 工具，取代了 DTS。
请务必记住，ETL 性能与您的架构以及您如何进行传输的关系与工具的关系一样重要。例如，如果您有索引，它们的运行速度会比您进行批量传输并在完成后创建索引时慢。如果您一次性执行大量操作，则会创建回滚日志，这些日志会增加大小并减慢进程速度。较小的批次可能会运行得更快，因为回滚日志不必那么大。
不要屈服于下意识的反应并责怪该工具。批判性地审视自己的做法，确保不会搬起石头砸自己的脚。
I believe SSIS is Microsoft's ETL tool today, replacing DTS.
It's important to remember that ETL performance has as much to do with your schema and how you're doing the transfer as it does the tool. For example, if you've got indexes they'll run slower than if you do a bulk transfer and create the indexes after it's done. If you do a large batch all at once you're creating rollback logs that increase in size and slow the process down. It could be that smaller batches will run faster, because the rollback log doesn't have to be as big.
Don't give in to the knee-jerk reaction and blame the tool. Look critically at how you're doing it to make sure that you're not shooting yourself in the foot.
没错，DTS 是 SSIS 之前用于 ETL 的 MS 工具。虽然我以前从未见过 DTS，但我相信与 DTS 相比，SSIS 更加用户友好且基于 GUI。说到用户友好性，我第一次接触 ETL 是在 Informatica，我坚信 Informatica 的用户友好性胜过 SSIS。 Inudstry 确实认识到 Informatica 比 SSIS 更加稳定和先进。
That's correct, DTS was MS tool for ETL prior to SSIS. While I have never seen DTS before, I believe SSIS is much more user friendly and GUI based in comparison to DTS. Speaking of user-friendly, my first experience with ETL was with Informatica, and I strongly believe that the user-friendliness of Informatica beats SSIS. Inudstry does recognize Informatica to be much more stable and advanced as opposed to SSIS.
SSIS 有它的问题
不能正确地与 Excel 一起工作（因为混合数据类型，众所周知的问题）
一切都在内存中执行 = 您需要大量内存。
特别是对于大文件的排序。
您无法指定使用哪种算法进行排序。
例如，如果能够使用合并排序就好了
因为不需要很多内存。
SSIS has got it's problems
Does not work with Excel correctly (because of mixed data types, well known problem)
Does everything in the memory = you need a lot of memory.
Especially for sorting large files.
You cannot specify which algorithm to use for sorting.
For example it would be nice to be able to use Merge sort
because does not require a lot of memory.
你的信息已经严重过时了。当前的 Microsoft ETL 工具是 SQL Server Integration Services (SSIS)。
Your information is badly out of date. The current Microsoft ETL tool is SQL Server Integration Services (SSIS).