dprevot/php-sqlsrv
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity
php-sqlsrv/test/Performance/performance_test_results.md
David Puglielli 316d5ac967 Update performance_test_results.md
2017-09-16 21:54:02 -04:00

15 KiB
Raw Blame History

Performance Test Results for the Microsoft Drivers for PHP for SQL Server

This page lists benchmarking results for both the SQLSRV and PDO_SQLSRV drivers for various operations and various environments. Test are performed on the following client environment:

  • PHP 7.1.7
  • Driver version 5.0.0
  • Platform: x64
  • Non-thread safe
  • ODBC Driver version 13.1

The client OS includes:

  • Windows Server 2016
  • MacOS Sierra
  • Red Hat 7.2
  • Ubuntu 16.04
  • SUSE 12

The server testing environment is either

  • Windows Server 2016 with SQL Server 2016, or
  • Ubuntu 16.04 with SQL Server Linux 2016

The following table lists benchmarking results for each scenario tested. Listed are the times taken for all iterations, and the maximum memory usage for all iterations in each scenario.

SQLSRV PDO_SQLSRV
Scenario Server Client Iterations Memory(MB) Time(s) Iterations Memory(MB) Time(s)
CRUD SQL Server Windows Windows Server 2016 1000x100 rows 0.88667 1074 1000x100 rows 0.88562 1195
OS X Sierra 1000x100 rows 0.88781 941 1000x100 rows 0.88678 1019
Red Hat 7.2 1000x100 rows 0.88223 1053 1000x100 rows 0.88120 1164
Ubuntu 16.04 1000x100 rows 0.88056 1059 1000x100 rows 0.87954 1172
Suse 12 1000x100 rows 0.88945 978 1000x100 rows 0.88849 1084
SQL Server Linux Windows Server 2016
OS X Sierra 1000x100 rows 0.88781 1112 1000x100 rows 0.88678 1210
Red Hat 7.2 1000x100 rows 0.88223 1189 1000x100 rows 0.88120 1315
Ubuntu 16.04 1000x100 rows 0.88056 1199
Suse 12
CRUD Insert SQL Server Windows Windows Server 2016 1000x100 rows 0.88380 311 1000x100 rows 0.88264 319
OS X Sierra 1000x100 rows 0.88770 280 1000x100 rows 0.88660 282
Red Hat 7.2 1000x100 rows 0.88342 304 1000x100 rows 0.88229 309
Ubuntu 16.04 1000x100 rows 0.88171 309 1000x100 rows 0.88065 357
Suse 12 1000x100 rows 0.88655 285 1000x100 rows 0.88548 290
SQL Server Linux Windows Server 2016 1000x100 rows 0.88681 368 1000x100 rows 0.88568 387
OS X Sierra 1000x100 rows 0.88770 348 1000x100 rows 0.88660 359
Red Hat 7.2 1000x100 rows 0.88342 349 1000x100 rows 0.88230 358
Ubuntu 16.04 1000x100 rows 0.88179 357 1000x100 rows 0.88067 363
Suse 12 1000x100 rows 0.88655 342 1000x100 rows 0.88548 344
CRUD Fetch SQL Server Windows Windows Server 2016 1000x100 rows 0.88431 255 1000x100 rows 0.88318 250
OS X Sierra 1000x100 rows 0.88828 188 1000x100 rows 0.88718 187
Red Hat 7.2 1000x100 rows 0.88396 248 1000x100 rows 0.88284 245
Ubuntu 16.04 1000x100 rows 0.88226 250 1000x100 rows 0.87514 247
Suse 12 1000x100 rows 0.88709 228 1000x100 rows 0.88602 227
SQL Server Linux Windows Server 2016 1000x100 rows 0.88735 287 1000x100 rows 0.88622 282
OS X Sierra 1000x100 rows 0.88828 229 1000x100 rows 0.88718 228
Red Hat 7.2 1000x100 rows 0.88400 272 1000x100 rows 0.88287 270
Ubuntu 16.04 1000x100 rows 0.88236 276 1000x100 rows 0.88124 274
Suse 12 1000x100 rows 0.88709 257 1000x100 rows 0.88602 254
CRUD Update SQL Server Windows Windows Server 2016 1000x100 rows 0.88594 314 1000x100 rows 0.88478 323
OS X Sierra 1000x100 rows 0.88988 284 1000x100 rows 0.88879 285
Red Hat 7.2 1000x100 rows 0.88560 307 1000x100 rows 0.88448 313
Ubuntu 16.04 1000x100 rows 0.87787 310 1000x100 rows 0.88235 327
Suse 12 1000x100 rows 0.88869 288 1000x100 rows 0.88762 294
SQL Server Linux Windows Server 2016 1000x100 rows 0.88899 369 1000x100 rows 0.88786 376
OS X Sierra 1000x100 rows 0.88988 349 1000x100 rows 0.88879 362
Red Hat 7.2 1000x100 rows 0.88560 357 1000x100 rows 0.88448 362
Ubuntu 16.04 1000x100 rows 0.88397 361 1000x100 rows 0.88285 367
Suse 12 1000x100 rows 0.88873 343 1000x100 rows 0.88766 340
CRUD Delete SQL Server Windows Windows Server 2016 1000x100 rows 0.88452 197 1000x100 rows 0.88341 303
OS X Sierra 1000x100 rows 0.88843 188 1000x100 rows 0.88737 274
Red Hat 7.2 1000x100 rows 0.88411 192 1000x100 rows 0.88303 295
Ubuntu 16.04 1000x100 rows 0.87638 193 1000x100 rows 0.88094 298
Suse 12 1000x100 rows 0.88728 181 1000x100 rows 0.88625 277
SQL Server Linux Windows Server 2016 1000x100 rows 0.88754 233 1000x100 rows 0.88645 358
OS X Sierra 1000x100 rows 0.88843 243 1000x100 rows 0.88737 340
Red Hat 7.2 1000x100 rows 0.88415 229 1000x100 rows 0.88307 342
Ubuntu 16.04 1000x100 rows 0.88251 229 1000x100 rows 0.88144 347
Suse 12 1000x100 rows 0.88728 220 1000x100 rows 0.88625 328
Large Read - 10,000,000 rows SQL Server Windows Windows Server 2016 1 0.88117 606 1 0.88003 382
OS X Sierra 1 0.88572 420 1 0.88463 344
Red Hat 7.2 1 0.88145 495 1 0.88035 694
Ubuntu 16.04 1 0.87981 413 1 0.87259 585
Suse 12 1 0.88397 453 1 0.88291 326
SQL Server Linux Windows Server 2016 1 0.88423 685
OS X Sierra 1 0.88571 308
Red Hat 7.2 1 0.88140 512
Ubuntu 16.04 1 0.87977 749
Suse 12 1 0.88397 828
SELECT @@Version SQL Server Windows Windows Server 2016 10,000 0.88074 16 10,000 0.87955 28
OS X Sierra 10,000 0.88522 19 10,000 0.88412 33
Red Hat 7.2 10,000 0.88118 11 10,000 0.88006 22
Ubuntu 16.04
Suse 12 10,000 0.88351 13 10,000 0.88243 24
SQL Server Linux Windows Server 2016 10,000 0.88377 16 10,000 0.88264 30
OS X Sierra 10,000 0.88522 19 10,000 0.88412 33
Red Hat 7.2 10,000 0.88094 17 10,000 0.87981 28
Ubuntu 16.04 10,000 0.87930 16 10,000 0.87818 26
Suse 12 10,000 0.88351 15 10,000 0.88243 26
CREATE DATABASE/TABLE/STORED PROCS SQL Server Windows Windows Server 2016 1000 0.88237 515 1000 0.88123 529
OS X Sierra 1000 0.88638 420 1000 0.88530 457
Red Hat 7.2 1000 0.88213 434 1000 0.88102 447
Ubuntu 16.04 1000 0.87437 395
Suse 12 1000 0.88512 494 1000 0.88406 507
SQL Server Linux Windows Server 2016 1000 0.88544 2285 1000 0.88433 2324
OS X Sierra 1000 0.88638 2338 1000 0.88530 2352
Red Hat 7.2 1000 0.88216 2292 1000 0.88106 2312
Ubuntu 16.04 1000 0.80530 2340 1000 0.87943 1228
Suse 12 1000 0.88512 2341 1000 0.88406 2337
Open and close 1000 connections SQL Server Windows Windows Server 2016 1000 0.87956 39 1000 0.87843 38
OS X Sierra 1000 0.88363 62 1000 0.88255 62
Red Hat 7.2 1000 0.87935 30 1000 0.87825 30
Ubuntu 16.04 1000 0.87608 29
Suse 12 1000 0.88232 24 1000 0.88126 25
SQL Server Linux Windows Server 2016 1000 0.88267 45 1000 0.88156 45
OS X Sierra 1000 0.88363 65 1000 0.88255 65
Red Hat 7.2 1000 0.87932 35 1000 0.87822 35
Ubuntu 16.04 1000 0.87771 35 1000 0.87659 34
Suse 12 1000 0.88232 29 1000 0.88126 30
Open and close 1000 connections with connection pooling SQL Server Windows Windows Server 2016
OS X Sierra
Red Hat 7.2
Ubuntu 16.04
Suse 12
SQL Server Linux Windows Server 2016
OS X Sierra
Red Hat 7.2
Ubuntu 16.04
Suse 12

The following table lists details of each scenario tested, including the SQL statements and data types used.

Scenario Description Iterations Operations/Iteration t-SQL Column Datatypes
CRUD Contains a loop for inserting a row into, fetching from, updating a row in, and deleting a row from a table. Measurement starts immediately before preparing the first INSERT INTO statement and ends immediately after the 100th DELETE statement executes. 1000 100
  • INSERT INTO <tableName> VALUES (<params>)
  • SELECT * FROM <tableName>
  • UPDATE <tableName> SET <params>
  • DELETE TOP(1) FROM <tableName>
  • VARCHAR(64)
  • NVARCHAR(64)
  • INT
  • DATETIME2
  • CHAR(64)
  • NCHAR(64)
  • NUMERIC
  • BINARY(64)
  • VARBINARY
  • DATETIMEOFFSET
CRUD Insert Contains a loop for inserting into a table. Each iteration prepares a statement, binds params, and executes. Measurement starts immediately before preparing the first INSERT INTO statement and ends immediately after the 100th statement executes. 1000 100 INSERT INTO <tableName> VALUES (<params>)
CRUD Fetch Contains a loop for fetching from a table with data inserted using the t-SQL in CRUD Insert. Each iteration prepares and executes a statement, and fetches a row from the result set. Measurement starts immediately before preparing the first SELECT statmeent and ends immediately after the 100th fetch. 1000 100 SELECT * FROM <tableName>
CRUD Update Contains a loop for updating a table with one row populated using the t-SQL in CRUD Insert. Each iteration prepares a statement, binds params, and executes. Measurement starts immediately before preparing the 1st UPDATE statement and ends immediately after the 100th statement executes. 1000 100 UPDATE <tableName> SET <params>
CRUD Delete Contains a loop for deleting one row from a table containing 100 rows which were populated uing the t-SQL in CRUD Insert. Each iteration prepares and executes a statement. Measurement starts immediately before preparing the 1st DELETE statement and ends immediately after the 100th statement executes. 1000 100 DELETE TOP(1) FROM <tableName>
Large Read Fetches one row at a time from a large prepopulated database until the whole result set is fetched. Measurement starts immediately before preparing the SELECT statement and ends immediately after the last fetch. 10,000,000 1 SELECT * FROM <tableName>
Select @Version Fetches the SQL Server version. Measurement starts immediately before executing the SELECT statement and ends immediately after fetch. 10,000 1 SELECT @@Version N/A
Create database, table, procedure Executes t-SQL statements to create a database, table, and procedure. Measurement starts immediately before executing the CREATE DATABASE statement and ends immediately after the DROP DATABASE statement. 1000 1
  • CREATE DATABASE <dbName>
  • USE <dbName>
  • CREATE TABLE <tableName> (<params>)
  • CREATE PROCEDURE <procName> @id INTEGER, @name VARCHAR(32) AS SET NOCOUNT ON; SELECT id, name, value FROM $databaseName.$tableName WHERE id = @id AND name = @name
  • USE MASTER; DROP DATABASE <dbName>
  • INT
  • VARCHAR(32)
  • INT
  • DATE
  • TIMESTAMP
  • TIME(7)
Connection Connects and disconnects from the database, with and without connection pooling. Measurement starts immediately before connecting and ends immediately after disconnecting. 1000 1 N/A