There was a new test/format failure. The likely cause was a recent optimization made to variable length column store traversal in WT-1833
nice catchsegv ./t runs=1 file_type=var data_source=file leaf_page_max=9 internal_page_max=9 key_min=10 value_min=256
t: process 11696
nextprev: prev key mismatch: 0000035561 != 35560
Configuration file:
############################################ # RUN PARAMETERS ############################################ abort=0 auto_throttle=1 firstfit=0 bitcnt=5 bloom=1 bloom_bit_count=20 bloom_hash_count=11 bloom_oldest=0 cache=33 checkpoints=1 checksum=uncompressed chunk_size=10 compaction=0 compression=bzip data_extend=1 data_source=file delete_pct=15 dictionary=0 evict_max=2 file_type=variable-length column-store backups=0 huffman_key=0 huffman_value=0 insert_pct=4 internal_key_truncation=1 internal_page_max=9 isolation=read-committed key_gap=15 key_max=93 key_min=10 leak_memory=0 leaf_page_max=9 logging=0 logging_archive=1 logging_prealloc=1 lsm_worker_threads=3 merge_max=19 mmap=1 ops=100000 prefix_compression=1 prefix_compression_min=3 repeat_data_pct=60 reverse=0 rows=100000 runs=1 split_pct=76 statistics=1 statistics_server=0 threads=1 timer=20 value_max=584 value_min=256 wiredtiger_config= write_pct=84 ############################################
A link to the original failure http://build.wiredtiger.com:8080/job/wiredtiger-test-format-stress/6523/console