NO CODE extract ZIP files with Pipelines in Microsoft Fabric

Рет қаралды 6,068

Күн бұрын

Пікірлер: 16

@jdg12800 10 ай бұрын

This is exactly what I needed to know! I had all those errors you warned about. Thank you for the very good step-by-step process. I am sure we will see improvements to the wizards soon.

@SameerShaikh-1992 10 ай бұрын

Very good and helpful video Adam. Thank you so much.

@JohnKpl 10 ай бұрын

Why did you choose Deflate over ZipDeflate?

@markkulofgren768 10 ай бұрын

Is there difference in performance doing it this way or coded way in the spirit of not being lazy but efficient?

@joebateman9668 10 ай бұрын

Great video Q. Can you explain what checking/unchecking enable staging does in the dataflow?

@rdeheld 10 ай бұрын

Wat would be the cheaper option if you do this in large amounts of data?

@GuyInACube 10 ай бұрын

Don't have a firm answer for you. It depends of course. What transformations are being done? What's the compute to handle that transformation. In general, my gut feel is the code/notebook option may be cheaper, but you should do some tests against your data to validate based on numbers.

@phenomenal325 10 ай бұрын

However great video lol, loved it thanks. Looks really similar to ADF

@Milhouse77BS 10 ай бұрын

Good video, but image frustration of business user trying to figure this out. :)

@phenomenal325 10 ай бұрын

You actually could have done all of that in just one step, I mean unless of course you wanted to move the zip file as well, but you could have just done the zip deflate and move the contents of the file in one activity rather than two.

@moeeljawad5361 7 ай бұрын

how can he do that? I did the equivalence of the two activities in one notebook function and i was thinking on how can we do it in pipeline solution. My data is very large, i could have a couple of hundreds of GB's in one folder.

@MrSparkefrostie 10 ай бұрын

Hmmm, zip file names have the same prefix, files inside are all data.csv

@DavidBridgeTechnology 5 ай бұрын

NO CODE is great until you need to make it resilient and multi-environment. It is hard to see what was created, so you can automate it or put it in source control, and if you use dataflow, you still cannot put these in source control at all (without hacky workaround using json extracts files). When will Microsoft wake up to the fact that source control should be property No1 on all features? Gen 1 dataflow was created in 2019 and still has not source control integration I'm not saying this is a bad video or a bad idea for people who need something quick and dirty, rather I am answering that question: Why do Deveopers always head for the notebook? The answer is: Because the process will be scripted and therefore fit nicely into the SDLC.