4paradigm / 4paradigm/OpenMLDB

disktable and memtable behave inconsistently with max_traverse_cnt

Open
#1,249 2 comments 0 reactions 1 assignee Claimed by @Leowner View on GitHub
bug storage-engine
Dominant language
C++
Stars
1.7k
Forks
331
Avg merge
12d 12h
Merged PRs (30d)
1

Description

When traversing a table, we use `max_traverse_cnt` to control. If `iterator->GetCount() >= max_traverse_cnt`, the traverse will be stopped. This decision is mainly in tablet_impl.cc but can be done early in disk_table.cc and mem_table.cc. For now, we add more `if (iterator->GetCount() >= max_traverse_cnt)` in disktable, which makes the `max_traverse_cnt` flag works in disktable but not in memtable. In consequence, some tests in table_test related to `max_traverse_cnt` have different results for memtable and disktable

Contributor guide

Open the contributing guide

Research direction

Look at tablet_impl.cc, disk_table.cc, and mem_table.cc for the traversal logic and the check on iterator->GetCount() >= max_traverse_cnt. The inconsistency is that disk_table has this check but mem_table does not. Run the table_test tests related to max_traverse_cnt to see the differing results. The fix is to ensure the same early stop logic is applied in both table implementations.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
databases
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.