-
Type: Bug
-
Resolution: Won't Fix
-
Priority: Minor - P4
-
None
-
Affects Version/s: None
-
Component/s: LSM
-
8
I saw LSM failures running test/format in the develop branch when truncate is configured, for example:
t: read_row: row 7: found in Berkeley DB, not found in WiredTiger t: read_row: row 4: found in Berkeley DB, not found in WiredTiger
WT_SESSION.truncate was just recently merged into format, and I'm pretty suspicious that's the issue. The CONFIG file is the same as WT-3927, which usually hangs, but I'm seeing it fail maybe 20% of the time.
The CONFIG file:
############################################ # RUN PARAMETERS ############################################ abort=0 alter=0 auto_throttle=1 backups=0 bitcnt=3 bloom=1 bloom_bit_count=15 bloom_hash_count=17 bloom_oldest=0 cache=300 cache_minimum=20 checkpoints=wiredtiger checkpoint_log_size=62 checkpoint_wait=85 checksum=on chunk_size=10 compaction=0 compression=snappy data_extend=0 data_source=lsm delete_pct=13 dictionary=0 direct_io=0 encryption=rotn-7 evict_max=3 file_type=row-store firstfit=0 huffman_key=0 huffman_value=0 independent_thread_rng=1 in_memory=0 insert_pct=61 internal_key_truncation=1 internal_page_max=9 isolation=random key_gap=14 key_max=256 key_min=256 leaf_page_max=9 leak_memory=0 logging=1 logging_archive=0 logging_compression=lz4-noraw logging_file_max=146059 logging_prealloc=0 long_running_txn=0 lsm_worker_threads=4 merge_max=17 mmap=1 modify_pct=19 ops=100000 prefix_compression=0 prefix_compression_min=0 quiet=1 read_pct=2 rebalance=1 repeat_data_pct=58 reverse=0 rows=100000 runs=1 salvage=1 split_pct=90 statistics=0 statistics_server=1 threads=1 timer=4 transaction_timestamps=0 transaction-frequency=49 truncate=1 value_max=100 value_min=32 verify=1 wiredtiger_config= write_pct=5 ############################################
- related to
-
WT-3927 Disable truncate testing with LSM
- Closed