Replies: 5 comments 9 replies
-
@MSBrett thoughts? |
Beta Was this translation helpful? Give feedback.
-
Hi @erincon01 The delay is due to how triggers fire in ADLS. In newer builds we no longer have a wait for the CSV files. The trigger fires when the manifest.json file is created, waits a bit for things to quiesce and then starts processing all the files in the manifest per the concurrency settings on the for loop. |
Beta Was this translation helpful? Give feedback.
-
Hi @MSBrett , I've deployed v0.2, and I see you have changed the trigger name to msexports_FileAdded. See that the trigger is disabled, and is configured to discard empty files. I am not sure if I've deployed the correct version (deploy from here: https://microsoft.github.io/finops-toolkit/hubs). { |
Beta Was this translation helpful? Give feedback.
-
I close this too. The trigger thing is fixed in the deploy you shared on friday. |
Beta Was this translation helpful? Give feedback.
-
I'd be happy to learn how to contribute. However, I am new in github :) Thanks. |
Beta Was this translation helpful? Give feedback.
-
The pipeline is wisely configured to wait 60 secs before processing the .csv files.
I have a few subscriptions where data exportation first creates an empty file, and after a few seconds the file is loaded.
Having the trigger to discard empty files, makes not to tranform the .csv to .parquet.
I have changed the trigger configuration to allow empty files.
It seems like when you builded the pipeline you had in mind that takes a while to export the data, but when building the deploy, the "allow empty files" optoin was not considered.
You might find this useful for next releases.
Beta Was this translation helpful? Give feedback.
All reactions