[doris] branch master updated: [fix](localgc) check gc only cache directory (#15238)

[pengxiangyu]

This is an automated email from the ASF dual-hosted git repository.

pengxiangyu pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris.git


The following commit(s) were added to refs/heads/master by this push:
     new 38530100d8 [fix](localgc) check gc only cache directory (#15238)
38530100d8 is described below

commit 38530100d812004d1d5069875ffa7188f31ab27b
Author: luozenglin <37725793+luozeng...@users.noreply.github.com>
AuthorDate: Fri Dec 23 10:40:55 2022 +0800

    [fix](localgc) check gc only cache directory (#15238)
---
 be/src/io/cache/file_cache_manager.cpp | 3 ++-
 be/src/olap/rowset/beta_rowset.cpp     | 9 +++++++++
 be/src/olap/rowset/beta_rowset.h       | 2 ++
 3 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/be/src/io/cache/file_cache_manager.cpp 
b/be/src/io/cache/file_cache_manager.cpp
index 6d34434190..a49dfcd240 100644
--- a/be/src/io/cache/file_cache_manager.cpp
+++ b/be/src/io/cache/file_cache_manager.cpp
@@ -22,6 +22,7 @@
 #include "io/cache/sub_file_cache.h"
 #include "io/cache/whole_file_cache.h"
 #include "io/fs/local_file_system.h"
+#include "olap/rowset/beta_rowset.h"
 #include "olap/storage_engine.h"
 #include "util/file_utils.h"
 #include "util/string_util.h"
@@ -125,7 +126,7 @@ void 
FileCacheManager::_gc_unused_file_caches(std::list<FileCachePtr>& result) {
             for (Path seg_file : seg_file_paths) {
                 std::string seg_filename = seg_file.native();
                 // check if it is a dir name
-                if (ends_with(seg_filename, ".dat") || ends_with(seg_filename, 
"clone")) {
+                if (!BetaRowset::is_segment_cache_dir(seg_filename)) {
                     continue;
                 }
                 // skip file cache already in memory
diff --git a/be/src/olap/rowset/beta_rowset.cpp 
b/be/src/olap/rowset/beta_rowset.cpp
index 8f0bda1bc6..a6a0d29870 100644
--- a/be/src/olap/rowset/beta_rowset.cpp
+++ b/be/src/olap/rowset/beta_rowset.cpp
@@ -58,6 +58,15 @@ std::string BetaRowset::segment_cache_path(int segment_id) {
     return fmt::format("{}/{}_{}", _tablet_path, rowset_id().to_string(), 
segment_id);
 }
 
+// just check that the format is xxx_segmentid and segmentid is numeric
+bool BetaRowset::is_segment_cache_dir(const std::string& cache_dir) {
+    auto segment_id_pos = cache_dir.find_last_of('_') + 1;
+    if (segment_id_pos >= cache_dir.size() || segment_id_pos == 0) {
+        return false;
+    }
+    return std::all_of(cache_dir.cbegin() + segment_id_pos, cache_dir.cend(), 
::isdigit);
+}
+
 std::string BetaRowset::segment_file_path(const std::string& rowset_dir, const 
RowsetId& rowset_id,
                                           int segment_id) {
     // {rowset_dir}/{schema_hash}/{rowset_id}_{seg_num}.dat
diff --git a/be/src/olap/rowset/beta_rowset.h b/be/src/olap/rowset/beta_rowset.h
index 8c43a742d2..a1167d2c2c 100644
--- a/be/src/olap/rowset/beta_rowset.h
+++ b/be/src/olap/rowset/beta_rowset.h
@@ -49,6 +49,8 @@ public:
     static std::string segment_cache_path(const std::string& rowset_dir, const 
RowsetId& rowset_id,
                                           int segment_id);
 
+    static bool is_segment_cache_dir(const std::string& cache_dir);
+
     static std::string segment_file_path(const std::string& rowset_dir, const 
RowsetId& rowset_id,
                                          int segment_id);
 


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org
For additional commands, e-mail: commits-h...@doris.apache.org