disk full and Elasticsearch won't run

[adamphetamine]

Hi All,
I have a silly chicken and egg situation.

VM disk is full, but instead of going into read only mode, it now won't launch.
This means I can't clean up the indices in the recommended way.
In addition, simply deleting the indices from disk fails- even with sudo.

The Docker repo is 55gb and seems to contain copies of both 2.3.250 and 2.3.260 images.
Deleting the older image set appeared to work but did not free any space.

Pruning the docker images appeared to free up 11gb but this extra space didn't actually appear either.

Yes, the obvious solution is to expand the VM disk size, but if possible I'd like to solve this by deleting some cruft and setting the curator correctly- any hints please?

/dev/vda2
111G nsm
34G var
2.3G usr
1.8G root
1.5G opt

/nsm
55G docker-registry
51G elasticsearch
2.5G influxdb
1.6G wazuh
1.3G repo

[adamphetamine]

gave up and expanded the disk...