EXPLAIN ANALYZE#
Примечание
Ниже приведена оригинальная документация Trino. Скоро мы ее переведем на русский язык и дополним полезными примерами.
Синтаксис#
EXPLAIN ANALYZE [VERBOSE] statement
Описание#
Execute the statement and show the distributed execution plan of the statement along with the cost of each operation.
The VERBOSE option will give more detailed information and low-level statistics;
understanding these may require knowledge of Trino internals and implementation details.
Примечание
The stats may not be entirely accurate, especially for queries that complete quickly.
Примеры#
In the example below, you can see the CPU time spent in each stage, as well as the relative cost of each plan node in the stage. Note that the relative cost of the plan nodes is based on wall time, which may or may not be correlated to CPU time. For each plan node you can see some additional statistics (e.g: average input per node instance). Such statistics are useful when one wants to detect data anomalies for a query (e.g: skewness).
EXPLAIN ANALYZE SELECT count(*), clerk FROM orders
WHERE orderdate > date '1995-01-01' GROUP BY clerk;
Query Plan
-----------------------------------------------------------------------------------------------
Trino version: version
Fragment 1 [HASH]
CPU: 22.58ms, Scheduled: 96.72ms, Blocked 46.21s (Input: 23.06s, Output: 0.00ns), Input: 1000 rows (37.11kB); per task: avg.: 1000.00 std.dev.: 0.00, Output: 1000 rows (28.32kB)
Output layout: [clerk, count]
Output partitioning: SINGLE []
Project[]
│ Layout: [clerk:varchar(15), count:bigint]
│ Estimates: {rows: ? (?), cpu: ?, memory: 0B, network: 0B}
│ CPU: 8.00ms (3.51%), Scheduled: 63.00ms (15.11%), Blocked: 0.00ns (0.00%), Output: 1000 rows (28.32kB)
│ Input avg.: 15.63 rows, Input std.dev.: 24.36%
└─ Aggregate[type = FINAL, keys = [clerk], hash = [$hashvalue]]
│ Layout: [clerk:varchar(15), $hashvalue:bigint, count:bigint]
│ Estimates: {rows: ? (?), cpu: ?, memory: ?, network: 0B}
│ CPU: 8.00ms (3.51%), Scheduled: 22.00ms (5.28%), Blocked: 0.00ns (0.00%), Output: 1000 rows (37.11kB)
│ Input avg.: 15.63 rows, Input std.dev.: 24.36%
│ count := count("count_0")
└─ LocalExchange[partitioning = HASH, hashColumn = [$hashvalue], arguments = ["clerk"]]
│ Layout: [clerk:varchar(15), count_0:bigint, $hashvalue:bigint]
│ Estimates: {rows: ? (?), cpu: ?, memory: 0B, network: 0B}
│ CPU: 2.00ms (0.88%), Scheduled: 4.00ms (0.96%), Blocked: 23.15s (50.10%), Output: 1000 rows (37.11kB)
│ Input avg.: 15.63 rows, Input std.dev.: 793.73%
└─ RemoteSource[sourceFragmentIds = [2]]
Layout: [clerk:varchar(15), count_0:bigint, $hashvalue_1:bigint]
CPU: 0.00ns (0.00%), Scheduled: 0.00ns (0.00%), Blocked: 23.06s (49.90%), Output: 1000 rows (37.11kB)
Input avg.: 15.63 rows, Input std.dev.: 793.73%
Fragment 2 [SOURCE]
CPU: 210.60ms, Scheduled: 327.92ms, Blocked 0.00ns (Input: 0.00ns, Output: 0.00ns), Input: 1500000 rows (18.17MB); per task: avg.: 1500000.00 std.dev.: 0.00, Output: 1000 rows (37.11kB)
Output layout: [clerk, count_0, $hashvalue_2]
Output partitioning: HASH [clerk][$hashvalue_2]
Aggregate[type = PARTIAL, keys = [clerk], hash = [$hashvalue_2]]
│ Layout: [clerk:varchar(15), $hashvalue_2:bigint, count_0:bigint]
│ CPU: 30.00ms (13.16%), Scheduled: 30.00ms (7.19%), Blocked: 0.00ns (0.00%), Output: 1000 rows (37.11kB)
│ Input avg.: 818058.00 rows, Input std.dev.: 0.00%
│ count_0 := count(*)
└─ ScanFilterProject[table = hive:sf1:orders, filterPredicate = ("orderdate" > DATE '1995-01-01')]
Layout: [clerk:varchar(15), $hashvalue_2:bigint]
Estimates: {rows: 1500000 (41.48MB), cpu: 35.76M, memory: 0B, network: 0B}/{rows: 816424 (22.58MB), cpu: 35.76M, memory: 0B, network: 0B}/{rows: 816424 (22.58MB), cpu: 22.58M, memory: 0B, network: 0B}
CPU: 180.00ms (78.95%), Scheduled: 298.00ms (71.46%), Blocked: 0.00ns (0.00%), Output: 818058 rows (12.98MB)
Input avg.: 1500000.00 rows, Input std.dev.: 0.00%
$hashvalue_2 := combine_hash(bigint '0', COALESCE("$operator$hash_code"("clerk"), 0))
clerk := clerk:varchar(15):REGULAR
orderdate := orderdate:date:REGULAR
Input: 1500000 rows (18.17MB), Filtered: 45.46%, Physical Input: 4.51MB
When the VERBOSE option is used, some operators may report additional information.
For example, the window function operator will output the following:
EXPLAIN ANALYZE VERBOSE SELECT count(clerk) OVER() FROM orders
WHERE orderdate > date '1995-01-01';
Query Plan
-----------------------------------------------------------------------------------------------
...
─ Window[]
│ Layout: [clerk:varchar(15), count:bigint]
│ CPU: 157.00ms (53.40%), Scheduled: 158.00ms (37.71%), Blocked: 0.00ns (0.00%), Output: 818058 rows (22.62MB)
│ metrics:
│ 'CPU time distribution (s)' = {count=1.00, p01=0.16, p05=0.16, p10=0.16, p25=0.16, p50=0.16, p75=0.16, p90=0.16, p95=0.16, p99=0.16, min=0.16, max=0.16}
│ 'Input rows distribution' = {count=1.00, p01=818058.00, p05=818058.00, p10=818058.00, p25=818058.00, p50=818058.00, p75=818058.00, p90=818058.00, p95=818058.00, p99=818058.00, min=818058.00, max=818058.00}
│ 'Scheduled time distribution (s)' = {count=1.00, p01=0.16, p05=0.16, p10=0.16, p25=0.16, p50=0.16, p75=0.16, p90=0.16, p95=0.16, p99=0.16, min=0.16, max=0.16}
│ Input avg.: 818058.00 rows, Input std.dev.: 0.00%
│ Active Drivers: [ 1 / 1 ]
│ Index size: std.dev.: 0.00 bytes, 0.00 rows
│ Index count per driver: std.dev.: 0.00
│ Rows per driver: std.dev.: 0.00
│ Size of partition: std.dev.: 0.00
│ count := count("clerk") RANGE UNBOUNDED_PRECEDING CURRENT_ROW
...