Use suffixed index names by default to prevent alias creation conflicts

[dmeremyanin]

Synopsis

Recently, we encountered a problem during index reset operations where a race condition could cause an alias creation failure. The issue arises when:

1. An index (e.g., users) is being reset.
2. The reset task creates a new suffixed index (e.g., users_<timestamp-suffix>) and starts the import process.
3. Once the import is finished, the reset task deletes the old index (here).
4. The reset task attempts to create an alias (e.g., users) pointing to the new suffixed index (here).
5. The problem occurs: There is a small time window between the old index deletion and the alias creation, during which another background job could recreate the unsuffixed users index. This results in a conflict when trying to create the alias, as both an index and an alias cannot share the same name.
6. The result: The alias creation fails, and no data is available in the index – the old index is deleted, and the new suffixed index is not aliased.

Solution

To prevent this issue, I updated the index creation logic to always use suffixed index names and unsuffixed aliases by default, not just after an index reset. This ensures that the alias conflict issue caused by the recreation of the old unsuffixed index is avoided.

---

Before submitting the PR make sure the following are checked:

• The PR relates to only one subject with a clear title and description in grammatically correct, complete sentences.
• Wrote good commit messages.
• Commit message starts with [Fix #issue-number] (if the related issue exists).
• Feature branch is up-to-date with master (if not - rebase it).
• Squashed related commits together.
• Added tests.
• Added an entry to the changelog if the new code introduces user-observable changes. See changelog entry format for details.

[dmeremyanin]

@toptal-anvil ping reviewers