[Magma - Volume Test]: Data ingestion started well but seems to be stuck when RAM is reduced. New dataset created is ingesting data properly.

1. Create a 3 KV node cluster.
2. Create a magma bucket with 2 replicas
3. Create 20 collections
4. Add 500M items in each collection. Total items = 10 billion
5. Start deleting the items.
6. While delete is going on add a cbas node to the cluster. CBAS RAM allocated is 256GB
7. Create few datasets on different collections.
I created:

Created analytics collection Default.ds7 on named collection GleamBookUsers0._default.VolumeCollection6

Created analytics collection Default.ds6 on named collection GleamBookUsers0._default.VolumeCollection5

Created analytics collection Default.ds5 on named collection GleamBookUsers0._default.VolumeCollection0

Created analytics collection Default.ds4 on named collection GleamBookUsers0._default.VolumeCollection3

Created analytics collection Default.ds3 on named collection GleamBookUsers0._default.VolumeCollection3

Created analytics collection Default.ds on named collection GleamBookUsers0._default.VolumeCollection2

Created analytics collection Default.ds2 on named collection GleamBookUsers0._default.VolumeCollection1

Created analytics collection Default.ds1 on named collection GleamBookUsers0._default.VolumeCollection0

8. While ingestion in ongoing after few minutes may be 30, change the RAM to 200GB.
9. Observe that after some time the ingestion is stuck.

10. Stop the delete load and at this time there are close to 8 billion items in the bucket overall.
11. Create another dataset ds8. Ingestion is working fine for this dataset. But, it is still stuck for all the remaining datasets.
12. Disconnect and connect link again also doesn't work.

Note: Total KV data on disk is close to 6TB.
I did not find any errors as such the the cbas logs.