why is recommended default setting is delta deleted file duration to be 7 days?

[Avvar2022]

Due to frequent updates on table our backend storage size is growing a lot, even though we have vacuum and optimize scheduled unable to clean up files 7 days or less.

Current settings is: delta.logRetentionDuration="interval 7 days" and deleted files setting is default (which is 7 days)

Based on the documentation, it is a recommended/default setting to keep  delta.deletedFileRetentionDuration="interval 7 days" and delta.logRetentionDuration="interval 7 days"but am not able to find a good documentation as why it is recommended to keep 7 days. what are the drawbacks if this setting is set to 24 hours

My scenario:

My table updates every 2-3 mins, I don't need delta log more than one day. i would like to run vacuum & optimize daily and clean up and consolidate files. 

Appreciate if you can share documentation or rational for having 7 days.

[szymon_dybczak]

Hi @Avvar2022 ,

I guess this recommendation is about cases where there was some kind of bug and you didn't notice it immediately, but after couple of days. Then you can restore table to previous version. So it's reasonable in such cases to have rcan retention period longer than one day 

But certainly, you can decrease log retention period if in your use case it's causing problems. It's totally valid option.

[Witold]

It's not really a recommendation per se, it's basically a default, which you simply need. And yes, it's supposed to be adapted to your specific needs.

In this case:

---

I don't need delta log more than one day.

---

If you're fine that you won't be able to rollback your data to previous days, then set it to one day, it's totally fine. You can even disable it completely, if you don't need this feature at all.