Improve Query Performance

performancequery-performancesqlite

I have the following sqlite3-Tables:

db.execute("""CREATE TABLE IF NOT EXISTS STATIONS
       (StationUID INTEGER PRIMARY KEY AUTOINCREMENT,
       StationNumber           TEXT    NOT NULL,
       StationName             TEXT    NOT NULL,
       StationLongitude        REAL    NOT NULL,
       StationAltitude         REAL    NOT NULL);""")

print "Created table STATIONS successfully"

db.execute("""CREATE TABLE IF NOT EXISTS TRAVELDAYS
       (TrainDaysUID           INT  NOT NULL,
       Day                     INT  NOT NULL,
       Value                   INT  NOT NULL,
       PRIMARY KEY (TrainDaysUID, Day));""")

print "Created table TRAVELDAYS successfully"

db.execute("""CREATE TABLE IF NOT EXISTS TRAINS
       (TrainUID INTEGER PRIMARY KEY AUTOINCREMENT,
       TrainNumber           TEXT    NOT NULL,
       TrainDaysUID          INT     NOT NULL,
       TrainType             TEXT    NOT NULL,
       TrainZone             TEXT    NOT NULL,
       FOREIGN KEY(TrainDaysUID) REFERENCES TRAVELDAYS(TrainDaysUID));""")

print "Created table TRAINS successfully"


db.execute("""CREATE TABLE IF NOT EXISTS ROUTES
       (RouteUID INTEGER PRIMARY KEY AUTOINCREMENT,
       StartStationUID       INT     NOT NULL,
       EndStationUID         INT     NOT NULL,
       FOREIGN KEY(StartStationUID) REFERENCES STATIONS(StationUID),
       FOREIGN KEY(EndStationUID) REFERENCES STATIONS(StationUID),
       UNIQUE (StartStationUID, EndStationUID) ON CONFLICT IGNORE);""")

print "Created table ROUTES successfully"

db.execute("""CREATE TABLE IF NOT EXISTS LEGS
       (LegUID INTEGER PRIMARY KEY AUTOINCREMENT,
       TrainUID            INT     NOT NULL,
       RouteUID            INT     NOT NULL,
       StartTime           INT     NOT NULL,
       EndTime             INT     NOT NULL,
       TravelTime          INT     NOT NULL,
       DwellTime           INT             ,
       FOREIGN KEY(TrainUID) REFERENCES TRAINS(TrainUID),
       FOREIGN KEY(RouteUID) REFERENCES ROUTES(RouteUID));""")

print "Created table LEGS successfully"

Table STATIONS has 10965 entries
Table ROUTES has 13026 entries
Table LEGS has 610593 entries
Table TRAINS has 44920 entries

I need to improve the performance of this query:

cr.execute("SELECT l.TrainUID, l.RouteUID, l.LegUID, l.StartTime, l.EndTime, l.TravelTime, l.DwellTime, s1.StationName, s2.StationName "
           "FROM LEGS AS l "
           "INNER JOIN ROUTES as r ON (r.RouteUID = l.RouteUID) "
           "INNER JOIN STATIONS AS s1 ON (r.StartStationUID = s1.StationUID) "
           "INNER JOIN STATIONS AS s2 ON (r.EndStationUID = s2.StationUID) "
           "WHERE l.StartTime >= 800 AND l.EndTime <= 1200 AND "
           "l.TrainUID IN ( "
                "SELECT DISTINCT l.TrainUID "
                "FROM LEGS AS l "
                "INNER JOIN ROUTES as r ON (r.RouteUID = l.RouteUID) "
                "INNER JOIN STATIONS AS s1 ON (r.StartStationUID = s1.StationUID) "
                "INNER JOIN STATIONS AS s2 ON (r.EndStationUID = s2.StationUID) "
                "INNER JOIN TRAINS AS t ON (t.TrainUID = l.TrainUID) "
                "INNER JOIN TRAVELDAYS as d ON (t.TrainDaysUID = d.TrainDaysUID) "
                "WHERE t.TrainDaysUID == d.TrainDaysUID AND d.Day == 300 AND d.Value == 1 AND (s1.StationNumber = 8000105 OR s2.StationNumber = 8000105)) ")

What can you recommend?

EDIT
I created now those indexes, but it is not really faster.

db.execute("""CREATE INDEX Idx1 ON TRAVELDAYS(TrainDaysUID, Day, Value);""")
db.execute("""CREATE INDEX Idx0 ON STATIONS(StationNumber);""")

Best Answer

Running EXPLAIN QUERY PLAN on your query gives this:

0|0|0|SCAN TABLE LEGS AS l
0|0|0|EXECUTE LIST SUBQUERY 1
1|0|5|SEARCH TABLE TRAVELDAYS AS d USING AUTOMATIC PARTIAL COVERING INDEX (Day=? AND Value=?)
1|1|4|SEARCH TABLE TRAINS AS t USING AUTOMATIC COVERING INDEX (TrainDaysUID=?)
1|2|0|SEARCH TABLE LEGS AS l USING AUTOMATIC COVERING INDEX (TrainUID=?)
1|3|1|SEARCH TABLE ROUTES AS r USING INTEGER PRIMARY KEY (rowid=?)
1|4|2|SEARCH TABLE STATIONS AS s1 USING INTEGER PRIMARY KEY (rowid=?)
1|5|3|SEARCH TABLE STATIONS AS s2 USING INTEGER PRIMARY KEY (rowid=?)
1|0|0|USE TEMP B-TREE FOR DISTINCT
0|1|1|SEARCH TABLE ROUTES AS r USING INTEGER PRIMARY KEY (rowid=?)
0|2|2|SEARCH TABLE STATIONS AS s1 USING INTEGER PRIMARY KEY (rowid=?)
0|3|3|SEARCH TABLE STATIONS AS s2 USING INTEGER PRIMARY KEY (rowid=?)

Any "AUTOMATIC" index is an index that SQLite thinks is worth creating temporarily; try this:

CREATE INDEX x1 ON TRAVELDAYS(Day, Value);
CREATE INDEX x2 ON TRAINS(TrainDaysUID);
CREATE INDEX x3 ON LEGS(TrainUID);

Related Solutions

Postgresql – Postgres multiple joins slow query, how to store default child record

You write:

Each customer can have multiple sites, but only one should be displayed in this list.

Yet, your query retrieves all rows. That would be a point to optimize. But you also do not define which site is to be picked.

Either way, it does not matter much here. Your EXPLAIN shows only 5026 rows for the site scan (5018 for the customer scan). So hardly any customer actually has more than one site. Did you ANALYZE your tables before running EXPLAIN?

From the numbers I see in your EXPLAIN, indexes will give you nothing for this query. Sequential table scans will be the fastest possible way. Half a second is rather slow for 5000 rows, though. Maybe your database needs some general performance tuning?

Maybe the query itself is faster, but "half a second" includes network transfer? EXPLAIN ANALYZE would tell us more.

If this query is your bottleneck, I would suggest you implement a materialized view.

After you provided more information I find that my diagnosis pretty much holds.

The query itself needs 27 ms. Not much of a problem there. "Half a second" was the kind of misunderstanding I had suspected. The slow part is the network transfer (plus ssh encoding / decoding, possibly rendering). You should only retrieve 100 rows, that would solve most of it, even if it means to execute the whole query every time.

If you go the route with a materialized view like I proposed you could add a serial number without gaps to the table plus index on it - by adding a column row_number() OVER (<your sort citeria here>) AS mv_id.

Then you can query:

SELECT *
FROM   materialized_view
WHERE  mv_id >= 2700
AND    mv_id <  2800;

This will perform very fast. LIMIT / OFFSET cannot compete, that needs to compute the whole table before it can sort and pick 100 rows.

pgAdmin timing

When you execute a query from the query tool, the message pane shows something like:

Total query runtime: 62 ms.

And the status line shows the same time. I quote pgAdmin help about that:

The status line will show how long the last query took to complete. If a dataset was returned, not only the elapsed time for server execution is displayed, but also the time to retrieve the data from the server to the Data Output page.

If you want to see the time on the server you need to use SQL EXPLAIN ANALYZE or the built in Shift + F7keyboard shortcut or Query -> Explain analyze. Then, at the bottom of the explain output you get something like this:

Total runtime: 0.269 ms

Mysql – Simple query is slow on 4M-rows table

In addition to @tombom's suggestions, creating an index on (user_id, post_id) instead of (or in addition, but the less indexes the better) separate indexes on user_id and post_id will simplify the query, probably getting rid of the filesort and temporary table, plus giving you the benefits of a covering index.

This will probably lower the query execution significantly if you have a large enough buffer pool and the query is relatively frequent.

If after doing that, the query is still slow, you will need to do (pre)caching in order to speed up the query execution.

Best Answer

Related Solutions

Postgresql – Postgres multiple joins slow query, how to store default child record

pgAdmin timing

Mysql – Simple query is slow on 4M-rows table

Related Question