Legacy Retention Strategies should not allow Max number of indices to be lower then 1

[kodjo-anipah]

Configuring the max number of indices to 0 leads to errors during retention.
Index Time & Index Message Count retention strategies are affected.

Expected Behavior

It should not be possible to set a value for the max number of indices to something lower than 1

Current Behavior

At some point, retention stops working at all since we always run into this exception:

2025-02-06 11:26:26,598 INFO : org.graylog2.indexer.retention.executors.CountBasedRetentionExecutor - Number of indices (2) higher than limit (0). Running retention for 2 indices.
2025-02-06 11:26:26,605 ERROR: org.graylog2.periodical.IndexRetentionThread - Uncaught exception in Periodical
java.lang.IllegalArgumentException: -1
	at java.base/java.util.stream.ReferencePipeline.skip(ReferencePipeline.java:575) ~[?:?]

Steps to Reproduce (for bugs)

1. Configure an index to Index Time Retention strategy and max number of indices to 0
2. Restart your Graylog server
3. See retention breaking