Changing retention rule in Router causes dropped data

Hi Folks,

I had a question about changing a retention rule on the Router (v0.15.1).

When I had set the rule from the default of loadForever to something else (loadByPeriod(P1Y), dropForever).

It now shows that all the segments became unavailable resulting in none of the data being able to be queried:

0.0% available (10,225 segments, 10,225 segments unavailable)

There was ~1.4TB of data, while that new rule should have removed ~500 GB. Is it expected to be unavailable initially and is there an estimate for when the data will become available?


Druid run these rules in the same order it is defined. So if dropForever is the last rule then it’ll run in the last and drop every segment.
If you do a dropForever, loadByPeriod(P1Y) then I think it should work.

Thanks for the response Naveen. It turns out that the replicationfactor was set to 0, causing the data to become available.