[improvement](session variable)Add enable_file_cache session variable (#16268)

Add enable_file_cache session variable, so that we can close file cache without restart BE.

be/src/io/file_factory.cpp

@@ -35,6 +35,7 @@
 #include "io/local_file_writer.h"
 #include "io/s3_reader.h"
 #include "io/s3_writer.h"
+#include "olap/iterators.h"
 #include "runtime/exec_env.h"
 #include "runtime/stream_load/load_stream_mgr.h"
 #include "runtime/stream_load/new_load_stream_mgr.h"
@@ -155,7 +156,7 @@ Status FileFactory::create_file_reader(RuntimeProfile* /*profile*/,
                                        io::FileReaderSPtr* file_reader, IOContext* io_ctx) {
     TFileType::type type = system_properties.system_type;
     auto cache_policy = io::FileCachePolicy::NO_CACHE;
-    if (config::enable_file_cache) {
+    if (config::enable_file_cache && io_ctx->enable_file_cache) {
         cache_policy = io::FileCachePolicy::FILE_BLOCK_CACHE;
     }
     io::FileReaderOptions reader_options(cache_policy, io::FileBlockCachePathPolicy());

be/src/olap/iterators.h

@@ -38,18 +38,20 @@ struct IOContext {
     IOContext() = default;
 
     IOContext(const TUniqueId* query_id_, FileCacheStatistics* stats_, bool is_presistent_,
-              bool use_disposable_cache_, bool read_segment_index_)
+              bool use_disposable_cache_, bool read_segment_index_, bool enable_file_cache)
             : query_id(query_id_),
               is_persistent(is_presistent_),
               use_disposable_cache(use_disposable_cache_),
               read_segment_index(read_segment_index_),
-              file_cache_stats(stats_) {}
+              file_cache_stats(stats_),
+              enable_file_cache(enable_file_cache) {}
     ReaderType reader_type;
     const TUniqueId* query_id = nullptr;
     bool is_persistent = false;
     bool use_disposable_cache = false;
     bool read_segment_index = false;
     FileCacheStatistics* file_cache_stats = nullptr;
+    bool enable_file_cache = true;
 };
 namespace vectorized {
 struct IteratorRowRef;

be/src/vec/exec/scan/vfile_scanner.cpp

@@ -77,6 +77,7 @@ Status VFileScanner::prepare(
     _io_ctx.reset(new IOContext());
     _io_ctx->file_cache_stats = _file_cache_statistics.get();
     _io_ctx->query_id = &_state->query_id();
+    _io_ctx->enable_file_cache = _state->query_options().enable_file_cache;
 
     if (vconjunct_ctx_ptr != nullptr) {
         // Copy vconjunct_ctx_ptr from scan node to this scanner's _vconjunct_ctx.

docs/en/docs/advanced/variables.md

@@ -567,3 +567,7 @@ Translated with www.DeepL.com/Translator (free version)
 *   `group_by_and_having_use_alias_first`
 
     Specifies whether group by and having clauses use column aliases rather than searching for column name in From clause. The default value is false.
+
+* `enable_file_cache`
+
+    Set wether to use block file cache. This variable takes effect only if the BE config enable_file_cache=true. The cache is not used when BE config enable_file_cache=false.

docs/zh-CN/docs/advanced/variables.md

@@ -555,4 +555,8 @@ SELECT /*+ SET_VAR(query_timeout = 1, enable_partition_cache=true) */ sleep(3);
 *   `group_by_and_having_use_alias_first`
 
        指定group by和having语句是否优先使用列的别名，而非从From语句里寻找列的名字。默认为false。
+
+* `enable_file_cache`
+
+    控制是否启用block file cache。该变量只有在be.conf中enable_file_cache=true时才有效，如果be.conf中enable_file_cache=false，则block file cache一直处于禁用状态。
 	

fe/fe-core/src/main/java/org/apache/doris/qe/SessionVariable.java

@@ -259,6 +259,7 @@ public class SessionVariable implements Serializable, Writable {
 
     public static final String ENABLE_TWO_PHASE_READ_OPT = "enable_two_phase_read_opt";
     public static final String TWO_PHASE_READ_OPT_LIMIT_THRESHOLD = "two_phase_read_opt_limit_threshold";
+    public static final String ENABLE_FILE_CACHE = "enable_file_cache";
 
     public static final String GROUP_BY_AND_HAVING_USE_ALIAS_FIRST = "group_by_and_having_use_alias_first";
 
@@ -686,6 +687,10 @@ public class SessionVariable implements Serializable, Writable {
     @VariableMgr.VarAttr(name = GROUP_BY_AND_HAVING_USE_ALIAS_FIRST)
     public boolean groupByAndHavingUseAliasFirst = false;
 
+    // Whether enable block file cache. Only take effect when BE config item enable_file_cache is true.
+    @VariableMgr.VarAttr(name = ENABLE_FILE_CACHE, needForward = true)
+    public boolean enableFileCache = true;
+
     // If this fe is in fuzzy mode, then will use initFuzzyModeVariables to generate some variables,
     // not the default value set in the code.
     public void initFuzzyModeVariables() {
@@ -1388,6 +1393,14 @@ public class SessionVariable implements Serializable, Writable {
         }
     }
 
+    public boolean isEnableFileCache() {
+        return enableFileCache;
+    }
+
+    public void setEnableFileCache(boolean enableFileCache) {
+        this.enableFileCache = enableFileCache;
+    }
+
     /**
      * Serialize to thrift object.
      * Used for rest api.
@@ -1450,6 +1463,8 @@ public class SessionVariable implements Serializable, Writable {
 
         tResult.setExternalSortBytesThreshold(externalSortBytesThreshold);
 
+        tResult.setEnableFileCache(enableFileCache);
+
         return tResult;
     }
 

gensrc/thrift/PaloInternalService.thrift

@@ -192,6 +192,8 @@ struct TQueryOptions {
   58: optional i64 external_sort_bytes_threshold = 0
 
   59: optional i32 partitioned_hash_agg_rows_threshold = 0
+
+  60: optional bool enable_file_cache = true
 }