pt-query-digest: 95th < min

[Federico]

I ran pt-query-digest on a slow log with default parameters. Stats for one of the queries:

# Query 2: 0.06 QPS, 0.27x concurrency, ID 0x4EFCFF71C99DA64D at byte 29219297
# This item is included in the report because it matches --limit.
# Scores: V/M = 0.01
# Time range: 2017-11-28 06:30:35 to 12:34:13
# Attribute pct total min max avg 95% stddev median
# ============ === ======= ======= ======= ======= ======= ======= =======
# Count 5 1348
# Exec time 14 5865s 4s 5s 4s 5s 220ms 4s
# Lock time 12 1s 413us 2ms 783us 1ms 188us 799us
# Rows sent 1 56.27k 0 140 42.75 130.47 36.06 34.95
# Rows examine 39 3.11G 2.37M 2.37M 2.37M 2.26M 0 2.26M
# Rows affecte 0 0 0 0 0 0 0 0
# Bytes sent 4 88.05M 18.29k 174.27k 66.89k 165.97k 40.51k 56.74k
# Query size 41 12.12M 9.21k 9.21k 9.21k 8.89k 0 8.89k

For Rows examined, you can see that 95% (2.26M) > min (2.37M). It also looks unlikely that median and 95% are the same - but of course it's possible.

I cannot share the slow log, but please let me know if I can provide more non reserved info.

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PT-1462