`resume-upgrade` returns `failed` although the charm is ready for upgrade

[phvalguima]

Occasionally, resume-upgrade is failing at the early steps of an upgrade, although the cluster shows itself as ready for upgrade status.

This has been observed multiple times in this CI run: https://github.com/canonical/opensearch-operator/pull/267

We need to further investigate what is causing this early failures.

[github-actions[bot]]

https://warthogs.atlassian.net/browse/DPE-4292

[phvalguima]

Closing this issue, as I believe it is related to #303