branch-4.0: [improve](log) Add segment file info when bitshuffle page corruption detected apache#60547 (apache#60689)

Cherry-picked from apache#60547

Co-authored-by: Luwei <814383175@qq.com>

Author: 2 people

be/src/olap/rowset/segment_v2/bitshuffle_page.h

@@ -282,13 +282,15 @@ inline Status parse_bit_shuffle_header(const Slice& data, size_t& num_elements,
     num_elements = decode_fixed32_le((const uint8_t*)&data[0]);
     compressed_size = decode_fixed32_le((const uint8_t*)&data[4]);
     num_element_after_padding = decode_fixed32_le((const uint8_t*)&data[8]);
+    size_of_element = decode_fixed32_le((const uint8_t*)&data[12]);
     if (num_element_after_padding != ALIGN_UP(num_elements, 8)) {
         return Status::InternalError(
                 "num of element information corrupted,"
-                " _num_element_after_padding:{}, _num_elements:{}",
-                num_element_after_padding, num_elements);
+                " _num_element_after_padding:{}, _num_elements:{}, expected_padding:{},"
+                " compressed_size:{}, size_of_element:{}, data_size:{}",
+                num_element_after_padding, num_elements, ALIGN_UP(num_elements, 8), compressed_size,
+                size_of_element, data.size);
     }
-    size_of_element = decode_fixed32_le((const uint8_t*)&data[12]);
     switch (size_of_element) {
     case 1:
     case 2:

be/src/olap/rowset/segment_v2/column_reader.cpp

@@ -1981,9 +1981,14 @@ Status FileColumnIterator::_read_data_page(const OrdinalPageIndexIterator& iter)
     RETURN_IF_ERROR(
             _reader->read_page(_opts, iter.page(), &handle, &page_body, &footer, _compress_codec));
     // parse data page
-    RETURN_IF_ERROR(ParsedPage::create(std::move(handle), page_body, footer.data_page_footer(),
-                                       _reader->encoding_info(), iter.page(), iter.page_index(),
-                                       &_page));
+    auto st = ParsedPage::create(std::move(handle), page_body, footer.data_page_footer(),
+                                 _reader->encoding_info(), iter.page(), iter.page_index(), &_page);
+    if (!st.ok()) {
+        LOG(WARNING) << "failed to create ParsedPage, file=" << _opts.file_reader->path().native()
+                     << ", page_offset=" << iter.page().offset << ", page_size=" << iter.page().size
+                     << ", page_index=" << iter.page_index() << ", error=" << st;
+        return st;
+    }
 
     // dictionary page is read when the first data page that uses it is read,
     // this is to optimize the memory usage: when there is no query on one column, we could

be/src/olap/rowset/segment_v2/indexed_column_reader.cpp

@@ -165,6 +165,11 @@ Status IndexedColumnIterator::_read_data_page(const PagePointer& pp) {
     opts.need_check_bitmap = false;
     status = ParsedPage::create(std::move(handle), body, footer.data_page_footer(),
                                 _reader->encoding_info(), pp, 0, &_data_page, opts);
+    if (!status.ok()) {
+        LOG(WARNING) << "failed to create ParsedPage in IndexedColumnIterator, file="
+                     << _reader->_file_reader->path().native() << ", page_offset=" << pp.offset
+                     << ", page_size=" << pp.size << ", error=" << status;
+    }
     DCHECK(_reader->_meta.ordinal_index_meta().is_root_data_page()
                    ? _reader->_meta.num_values() == _data_page.num_rows
                    : true);