More read-only-allow-delete docs (#45320)

Adds to the `index.blocks.read_only_allow_delete` docs the information that
this block may be added or removed automatically, and rewords the
breaking-changes docs to mention the blocks explicitly and to recommend using a
different block.

Relates #42559

docs/reference/index-modules.asciidoc

@@ -166,8 +166,10 @@ specific index module:
     allow writes and metadata changes.
 
 `index.blocks.read_only_allow_delete`::
-    Identical to `index.blocks.read_only` but allows deleting the index to free
+
-     up resources.
+    Similar to `index.blocks.read_only` but also allows deleting the index to
+    free up resources. The <<disk-allocator,disk-based shard allocator>> may
+    add and remove this block automatically.
 
 `index.blocks.read`::
 

docs/reference/migration/migrate_7_4.asciidoc

@@ -99,14 +99,17 @@ wait for the rerouting process to completely finish you should add the
 [float]
 ==== Auto-release of read-only-allow-delete block
 
-If a node exceeds the flood-stage disk watermark then we add a block to all of
+If a node exceeds the flood-stage disk watermark then {es} adds the
-its indices to prevent further writes as a last-ditch attempt to prevent the
+`index.blocks.read_only_allow_delete` block to all of its indices to prevent
-node completely exhausting its disk space. In earlier versions this block would
+further writes, as a last-resort attempt to prevent the node completely
-remain in place until manually removed, causing confusion for users who
+exhausting its disk space. In earlier versions this block would remain in place
-currently have ample disk space and are not aware that they nearly ran out at
+until manually removed, causing confusion for users who currently have ample
-some point in the past. From 7.4 onwards the block is automatically removed
+disk space and who are not aware that they nearly ran out at some point in the
-when a node drops below the high watermark again, with the expectation that the
+past. From 7.4 onwards the block is automatically removed when the node drops
-high watermark is some distance below the flood-stage watermark and therefore
+below the high watermark again, with the expectation that the high watermark is
-the disk space problem is truly resolved. This behaviour can be disabled by
+some distance below the flood-stage watermark and therefore the disk space
-setting the system property `es.disk.auto_release_flood_stage_block` to
+problem is truly resolved. Since this block may be automatically removed, you
-`false`.
+can no longer rely on adding this block manually to prevent writes to an index.
+You should use the `index.blocks.read_only` block instead. This behaviour can
+be disabled by setting the system property
+`es.disk.auto_release_flood_stage_block` to `false`.

docs/reference/modules/cluster/disk_allocator.asciidoc

@@ -42,8 +42,6 @@ shards allocated on the node that has at least one disk exceeding the flood
 stage. This is a last resort to prevent nodes from running out of disk space.
 The index block is automatically released once the disk utilization falls below
 the high watermark.
-The automatic release can however be disabled in 7.x through a system property
-`es.disk.auto_release_flood_stage_block`
 
 NOTE: You can not mix the usage of percentage values and byte values within
 these settings. Either all are set to percentage values, or all are set to byte