hi all,
there may be a bug which is caused by loading duplicate rows to a unique
table 。

be log:
start cumulative compaction.
tablet=1330586.122703930.2443b3861700c0e6-9e12fa77ec8224b5,
output_version=[2-8], permits: 7
row_num does not match between cumulative input and output!
input_row_num=1000002, merged_row_num=2, filtered_row_num=0,
output_row_num=674368


the tablet info:
"rowsets": [
        "[0-1] 0 DATA NONOVERLAPPING
02000000000017cefb46ec3596ad71f03b06b495e6ba1097 0",
        "[2-6] 1 DATA NONOVERLAPPING
02000000000017d4fb46ec3596ad71f03b06b495e6ba1097 2.92 KB", 1 row
        "[7-7] 1 DATA NONOVERLAPPING
02000000000017d5fb46ec3596ad71f03b06b495e6ba1097 2.92 KB", 1 row
        "[8-8] 5 DATA OVERLAPPING
02000000000017d6fb46ec3596ad71f03b06b495e6ba1097 39.99 MB" 1000000 row
    ],
    "missing_rowsets": [],
    "stale_rowsets": [
        "[2-2] 1 02000000000017cffb46ec3596ad71f03b06b495e6ba1097 2.92 KB",
        "[3-3] 1 02000000000017d0fb46ec3596ad71f03b06b495e6ba1097 2.92 KB",
        "[4-4] 1 02000000000017d1fb46ec3596ad71f03b06b495e6ba1097 2.92 KB",
        "[5-5] 1 02000000000017d2fb46ec3596ad71f03b06b495e6ba1097 2.92 KB",
        "[6-6] 1 02000000000017d3fb46ec3596ad71f03b06b495e6ba1097 2.92 KB"
    ],

Reply via email to