Add 11-25 relese notes

Author: chenjian2664

docs/src/main/sphinx/release/release-479.md

@@ -10,6 +10,8 @@
 * Require the JDK Vector API to be enabled at runtime ({issue}`27340`)
 * Add support for `SET DEFAULT` and `DROP DEFAULT` statements. ({issue}`26162`)
 * Add `array_first` and `array_last` functions. ({issue}`27295`)
+* Deprecate `EXPLAIN (TYPE LOGICAL)`. `EXPLAIN (TYPE DISTRIBUTED)` should be used instead. ({issue}`27434`)
+* Improve performance of data exchanges by using SIMD instructions on ARM CPUs that support SVE, such as graviton 3 and 4. This can be disabled by setting `experimental.blockserde-vectorized-null-suppression-strategy=NONE`({issue}`27426`)
 * Enhance after `RemoveTrivialFilters` call `RemoveEmptyUnionBranches` to prune empty union branches. ({issue}`21506`)
 * Improve performance of {func}`array_sort` function. ({issue}`27272`)
 * Improve performance of `repeat` function. ({issue}`27369`)
@@ -51,6 +53,8 @@
 * Harden hierarchical namespace check in Azure with root blob fallback check. ({issue}`27278`)
 * Avoid reading unusually large pages from parquet files.({issue}`27148`)
 * Fix potential `FileAlreadyExistsException` failure when network failures during s3 write({issue}`27330`)
+* Prevent log writer from failing with FileAlreadyExistsException when there are network failures during write. 
+  writing to s3 now requires permissions for PutObjectTagging and GetObjectTagging operations (breaking change).({issue}`27388`)
 
 ## Druid connector