第一步：
配置maven
<dependency>
<groupid>p6spy</groupid>
<artifactid>p6spy</artifactid>
<version>3.6.0</version>
</dependency>
<dependency>
<groupid>com.alibaba</groupid>
<artifactid>druid</artifactid>
<version>1.1.9</version>
</dependency>
第二步配置数据源：
application.properties配置
#jdbc.driver=oracle.jdbc.driver.OracleDriver
#jdbc.driver=com.mysql.jdbc.Driver
jdbc.driver=com.p6spy.engine.spy.P6SpyDriver
#jdbc.url=jdbc:oracle:thin:@192.168.10.10:1521:orcl
#jdbc.url=jdbc:mysql://localhost:3306/dbname
jdbc.url=jdbc:p6spy:mysql://127.0.0.1:3306/mybatis?useUnicode=true&characterEncoding=UTF-8&useSSL=false
jdbc.username=root
jdbc.password=root
(ps:如果是oracle数据库将mysql换成oracle即可)
applicationContext.xml配置
<bean id="dataSourceTarget" class="com.mchange.v2.c3p0.ComboPooledDataSource">
<property name="driverClass" value="${jdbc.driver}" />
<property name="jdbcUrl" value="${jdbc.url}" />
<property name="user" value="${jdbc.username}" />
<property name="password" value="${jdbc.password}" />
<property name="initialPoolSize" value="${connection_pools.initial_pool_size}" />
<property name="minPoolSize" value="${connection_pools.min_pool_size}" />
<property name="maxPoolSize" value="${connection_pools.max_pool_size}" />
<property name="maxIdleTime" value="${connection_pools.max_idle_time}" />
<property name="acquireIncrement" value="${connection_pools.acquire_increment}" />
<property name="checkoutTimeout" value="${connection_pools.checkout_timeout}" />
</bean>
<bean id="dataSource" class="com.p6spy.engine.spy.P6DataSource">
<constructor-arg ref="dataSourceTarget"></constructor-arg>
</bean>
有部分同学将数据源的驱动包换成了com.p6spy.engine.spy.P6DataSource
其实不然，p6spy只是代理，其Driver并不能作为数据源的驱动包,故我们使用源Driver即可，在其基础上代理一层即可
第三步：
sql输出工具类P6SpyLogger
import java.util.Properties;
import org.apache.log4j.Logger;
import com.p6spy.engine.logging.Category;
import com.p6spy.engine.spy.appender.StdoutLogger;
import com.ruiec.web.util.RuiecStringUtil;
public class P6SpyLogger extends StdoutLogger {
protected final Logger logger = Logger.getLogger(this.getClass());
/**
* 重写输出方法
* @param connectionId 连接id
* @param now 当前时间
* @param elapsed 执行时长，包括执行 SQL 和处理结果集的时间(可以参考来调优)
* @param category 语句分类，statement、resultset 等
* @param prepared 查询语句。可能是 prepared statement，表现为 select * from table1 where c1=?，问号参数形式
* @param sql 含参数值的查询语句，如 select * from from table1 where c1=7
*/
@Override
public void logSQL(int connectionId, String now, long elapsed, Category category, String prepared, String sql) {
Properties properties = new Properties();
Boolean flag = false;
try {
properties.load(P6SpyLogger.class.getClassLoader().getResourceAsStream("application.properties"));
flag = Boolean.valueOf(properties.getProperty("sqlLogger"));
} catch (Exception e) {
flag = false;
}
if (flag) {
if(!Category.COMMIT.equals(category) && !prepared.startsWith("select count(")){
this.logText(this.strategy.formatMessage(connectionId, now, elapsed, category.toString(), "-prepared-", sql));
}
}
}
@Override
public void logText(String text) {
StringBuilder sb = new StringBuilder();
//匹配到最后一个|作为分隔符
String[] arrString = text.split("\\|(?![^\\|]*\\|)");
if (arrString.length > 1) {
sb.append(arrString[0]);
//去最后一段语句做替换进行格式化
String sss=arrString[1].trim();
if(!RuiecStringUtil.checkNull(sss) && !";".equalsIgnoreCase(sss)){
String sql = new SQLFormatter().format(arrString[1]);
sb.append("\r\n");
sb.append(sql);
sb.append("\r\n");
}else {
sb.append(sss);
}
//this.getStream().println(sb.toString());
logger.info(sb.toString());
} else {
//this.getStream().println(text);
logger.info(text);
}
arrString = null;
}
}
我做了一个开关在application.properties中,sqlLogger等于true时输出sql方便调试，不需要使用时设置为false则不输出sql
sql格式化工具类SQLFormatter
import java.util.HashSet;
import java.util.LinkedList;
import java.util.Set;
import java.util.StringTokenizer;
public class SQLFormatter {
private static final Set<String> BEGIN_CLAUSES = new HashSet<String>();
private static final Set<String> END_CLAUSES = new HashSet<String>();
private static final Set<String> LOGICAL = new HashSet<String>();
private static final Set<String> QUANTIFIERS = new HashSet<String>();
private static final Set<String> DML = new HashSet<String>();
private static final Set<String> MISC = new HashSet<String>();
public static final String WHITESPACE = " \n\r\f\t";
static {
BEGIN_CLAUSES.add("left");
BEGIN_CLAUSES.add("right");
BEGIN_CLAUSES.add("inner");
BEGIN_CLAUSES.add("outer");
BEGIN_CLAUSES.add("group");
BEGIN_CLAUSES.add("order");
END_CLAUSES.add("where");
END_CLAUSES.add("set");
END_CLAUSES.add("having");
END_CLAUSES.add("join");
END_CLAUSES.add("from");
END_CLAUSES.add("by");
END_CLAUSES.add("join");
END_CLAUSES.add("into");
END_CLAUSES.add("union");
LOGICAL.add("and");
LOGICAL.add("or");
LOGICAL.add("when");
LOGICAL.add("else");
LOGICAL.add("end");
QUANTIFIERS.add("in");
QUANTIFIERS.add("all");
QUANTIFIERS.add("exists");
QUANTIFIERS.add("some");
QUANTIFIERS.add("any");
DML.add("insert");
DML.add("update");
DML.add("delete");
MISC.add("select");
MISC.add("on");
}
static final String indentString = " ";
static final String initial = "\n ";
public String format(String source) {
return new FormatProcess(source).perform();
}
private static class FormatProcess {
boolean beginLine = true;
boolean afterBeginBeforeEnd = false;
boolean afterByOrSetOrFromOrSelect = false;
boolean afterValues = false;
boolean afterOn = false;
boolean afterBetween = false;
boolean afterInsert = false;
int inFunction = 0;
int parensSinceSelect = 0;
private LinkedList<Integer> parenCounts = new LinkedList<Integer>();
private LinkedList<Boolean> afterByOrFromOrSelects = new LinkedList<Boolean>();
int indent = 1;
StringBuilder result = new StringBuilder();
StringTokenizer tokens;
String lastToken;
String token;
String lcToken;
public FormatProcess(String sql) {
tokens = new StringTokenizer(
sql,
"()+*/-=<>'`\"[]," + WHITESPACE,
true
);
}
public String perform() {
result.append(initial);
while (tokens.hasMoreTokens()) {
token = tokens.nextToken();
lcToken = token.toLowerCase();
if ("'".equals(token)) {
String t;
do {
t = tokens.nextToken();
token += t;
}
while (!"'".equals(t) && tokens.hasMoreTokens()); // cannot handle single quotes
} else if ("\"".equals(token)) {
String t;
do {
t = tokens.nextToken();
token += t;
}
while (!"\"".equals(t));
}
if (afterByOrSetOrFromOrSelect && ",".equals(token)) {
commaAfterByOrFromOrSelect();
} else if (afterOn && ",".equals(token)) {
commaAfterOn();
} else if ("(".equals(token)) {
openParen();
} else if (")".equals(token)) {
closeParen();
} else if (BEGIN_CLAUSES.contains(lcToken)) {
beginNewClause();
} else if (END_CLAUSES.contains(lcToken)) {
endNewClause();
} else if ("select".equals(lcToken)) {
select();
} else if (DML.contains(lcToken)) {
updateOrInsertOrDelete();
} else if ("values".equals(lcToken)) {
values();
} else if ("on".equals(lcToken)) {
on();
} else if (afterBetween && lcToken.equals("and")) {
misc();
afterBetween = false;
} else if (LOGICAL.contains(lcToken)) {
logical();
} else if (isWhitespace(token)) {
white();
} else {
misc();
}
if (!isWhitespace(token)) {
lastToken = lcToken;
}
}
return result.toString();
}
private void commaAfterOn() {
out();
indent--;
newline();
afterOn = false;
afterByOrSetOrFromOrSelect = true;
}
private void commaAfterByOrFromOrSelect() {
out();
newline();
}
private void logical() {
if ("end".equals(lcToken)) {
indent--;
}
newline();
out();
beginLine = false;
}
private void on() {
indent++;
afterOn = true;
newline();
out();
beginLine = false;
}
private void misc() {
out();
if ("between".equals(lcToken)) {
afterBetween = true;
}
if (afterInsert) {
newline();
afterInsert = false;
} else {
beginLine = false;
if ("case".equals(lcToken)) {
indent++;
}
}
}
private void white() {
if (!beginLine) {
result.append(" ");
}
}
private void updateOrInsertOrDelete() {
out();
indent++;
beginLine = false;
if ("update".equals(lcToken)) {
newline();
}
if ("insert".equals(lcToken)) {
afterInsert = true;
}
}
@SuppressWarnings({"UnnecessaryBoxing"})
private void select() {
out();
indent++;
newline();
parenCounts.addLast(Integer.valueOf(parensSinceSelect));
afterByOrFromOrSelects.addLast(Boolean.valueOf(afterByOrSetOrFromOrSelect));
parensSinceSelect = 0;
afterByOrSetOrFromOrSelect = true;
}
private void out() {
result.append(token);
}
private void endNewClause() {
if (!afterBeginBeforeEnd) {
indent--;
if (afterOn) {
indent--;
afterOn = false;
}
newline();
}
out();
if (!"union".equals(lcToken)) {
indent++;
}
newline();
afterBeginBeforeEnd = false;
afterByOrSetOrFromOrSelect = "by".equals(lcToken)
|| "set".equals(lcToken)
|| "from".equals(lcToken);
}
private void beginNewClause() {
if (!afterBeginBeforeEnd) {
if (afterOn) {
indent--;
afterOn = false;
}
indent--;
newline();
}
out();
beginLine = false;
afterBeginBeforeEnd = true;
}
private void values() {
indent--;
newline();
out();
indent++;
newline();
afterValues = true;
}
@SuppressWarnings({"UnnecessaryUnboxing"})
private void closeParen() {
parensSinceSelect--;
if (parensSinceSelect < 0) {
indent--;
parensSinceSelect = parenCounts.removeLast().intValue();
afterByOrSetOrFromOrSelect = afterByOrFromOrSelects.removeLast().booleanValue();
}
if (inFunction > 0) {
inFunction--;
out();
} else {
if (!afterByOrSetOrFromOrSelect) {
indent--;
newline();
}
out();
}
beginLine = false;
}
private void openParen() {
if (isFunctionName(lastToken) || inFunction > 0) {
inFunction++;
}
beginLine = false;
if (inFunction > 0) {
out();
} else {
out();
if (!afterByOrSetOrFromOrSelect) {
indent++;
newline();
beginLine = true;
}
}
parensSinceSelect++;
}
private static boolean isFunctionName(String tok) {
final char begin = tok.charAt(0);
final boolean isIdentifier = Character.isJavaIdentifierStart(begin) || '"' == begin;
return isIdentifier &&
!LOGICAL.contains(tok) &&
!END_CLAUSES.contains(tok) &&
!QUANTIFIERS.contains(tok) &&
!DML.contains(tok) &&
!MISC.contains(tok);
}
private static boolean isWhitespace(String token) {
return WHITESPACE.indexOf(token) >= 0;
}
private void newline() {
result.append("\n");
for (int i = 0; i < indent; i++) {
result.append(indentString);
}
beginLine = true;
}
}
public static void main(String[] args) {
String sql = new SQLFormatter().format("select * from t_sss");
System.out.println(sql);
}
}
格式化工具类是为了使输出的sql更方便可读性，如下：
第四步：
配置 spy.properties
################################################################# # P6Spy Options File # # See documentation for detailed instructions # # http://p6spy.github.io/p6spy/2.0/configandusage.html # ################################################################# ################################################################# # MODULES # # # # Module list adapts the modular functionality of P6Spy. # # Only modules listed are active. # # (default is com.p6spy.engine.logging.P6LogFactory and # # com.p6spy.engine.spy.P6SpyFactory) # # Please note that the core module (P6SpyFactory) can't be # # deactivated. # # Unlike the other properties, activation of the changes on # # this one requires reload. # ################################################################# #modulelist=com.p6spy.engine.spy.P6SpyFactory,com.p6spy.engine.logging.P6LogFactory,com.p6spy.engine.outage.P6OutageFactory ################################################################ # CORE (P6SPY) PROPERTIES # ################################################################ # A comma separated list of JDBC drivers to load and register. # (default is empty) # # Note: This is normally only needed when using P6Spy in an # application server environment with a JNDI data source or when # using a JDBC driver that does not implement the JDBC 4.0 API # (specifically automatic registration). driverlist=com.mysql.jdbc.Driver # for flushing per statement # (default is false) #autoflush = false # sets the date format using Java's SimpleDateFormat routine. # In case property is not set, miliseconds since 1.1.1970 (unix time) is used (default is empty) dateformat=yyyy-MM-dd HH:mm:ss # prints a stack trace for every statement logged #stacktrace=false # if stacktrace=true, specifies the stack trace to print #stacktraceclass= # determines if property file should be reloaded # Please note: reload means forgetting all the previously set # settings (even those set during runtime - via JMX) # and starting with the clean table # (default is false) #reloadproperties=false # determines how often should be reloaded in seconds # (default is 60) #reloadpropertiesinterval=60 # specifies the appender to use for logging # Please note: reload means forgetting all the previously set # settings (even those set during runtime - via JMX) # and starting with the clean table # (only the properties read from the configuration file) # (default is com.p6spy.engine.spy.appender.FileLogger) #appender=com.p6spy.engine.spy.appender.Slf4JLogger #appender=com.p6spy.engine.spy.appender.StdoutLogger #appender=com.p6spy.engine.spy.appender.FileLogger #自定义的SQL格式化输出，换成文件存放包名 appender=cn.wuxia.ueq.ucm.view.base.support.P6SpyLogger # name of logfile to use, note Windows users should make sure to use forward slashes in their pathname (e:/test/spy.log) # (used for com.p6spy.engine.spy.appender.FileLogger only) # (default is spy.log) #logfile = spy.log # append to the p6spy log file. if this is set to false the # log file is truncated every time. (file logger only) # (default is true) #append=true # class to use for formatting log messages (default is: com.p6spy.engine.spy.appender.SingleLineFormat) logMessageFormat=com.p6spy.engine.spy.appender.SingleLineFormat #logMessageFormat=com.p6spy.engine.spy.appender.MultiLineFormat # format that is used for logging of the date/time/... (has to be compatible with java.text.SimpleDateFormat) # (default is dd-MMM-yy) databaseDialectDateFormat=yyyy-MM-dd HH:mm:ss # whether to expose options via JMX or not # (default is true) #jmx=true # if exposing options via jmx (see option: jmx), what should be the prefix used? # jmx naming pattern constructed is: com.p6spy(.<jmxPrefix>)?:name=<optionsClassName> # please note, if there is already such a name in use it would be unregistered first (the last registered wins) # (default is none) #jmxPrefix= # if set to true, the execution time will be measured in nanoseconds as opposed to milliseconds # (default is false) #useNanoTime=false ################################################################# # DataSource replacement # # # # Replace the real DataSource class in your application server # # configuration with the name com.p6spy.engine.spy.P6DataSource # # (that provides also connection pooling and xa support). # # then add the JNDI name and class name of the real # # DataSource here # # # # Values set in this item cannot be reloaded using the # # reloadproperties variable. Once it is loaded, it remains # # in memory until the application is restarted. # # # ################################################################# #realdatasource=/RealMySqlDS #realdatasourceclass=com.mysql.jdbc.jdbc2.optional.MysqlDataSource ################################################################# # DataSource properties # # # # If you are using the DataSource support to intercept calls # # to a DataSource that requires properties for proper setup, # # define those properties here. Use name value pairs, separate # # the name and value with a semicolon, and separate the # # pairs with commas. # # # # The example shown here is for mysql # # # ################################################################# #realdatasourceproperties=port;3306,serverName;myhost,databaseName;jbossdb,foo;bar ################################################################# # JNDI DataSource lookup # # # # If you are using the DataSource support outside of an app # # server, you will probably need to define the JNDI Context # # environment. # # # # If the P6Spy code will be executing inside an app server then # # do not use these properties, and the DataSource lookup will # # use the naming context defined by the app server. # # # # The two standard elements of the naming environment are # # jndicontextfactory and jndicontextproviderurl. If you need # # additional elements, use the jndicontextcustom property. # # You can define multiple properties in jndicontextcustom, # # in name value pairs. Separate the name and value with a # # semicolon, and separate the pairs with commas. # # # # The example shown here is for a standalone program running on # # a machine that is also running JBoss, so the JDNI context # # is configured for JBoss (3.0.4). # # # # (by default all these are empty) # ################################################################# #jndicontextfactory=org.jnp.interfaces.NamingContextFactory #jndicontextproviderurl=localhost:1099 #jndicontextcustom=java.naming.factory.url.pkgs;org.jboss.nameing:org.jnp.interfaces #jndicontextfactory=com.ibm.websphere.naming.WsnInitialContextFactory #jndicontextproviderurl=iiop://localhost:900 ################################################################ # P6 LOGGING SPECIFIC PROPERTIES # ################################################################ # filter what is logged # please note this is a precondition for usage of: include/exclude/sqlexpression # (default is false) #filter=false # comma separated list of strings to include # please note that special characters escaping (used in java) has to be done for the provided regular expression # (default is empty) #include = # comma separated list of strings to exclude # (default is empty) #exclude = # sql expression to evaluate if using regex # please note that special characters escaping (used in java) has to be done for the provided regular expression # (default is empty) #sqlexpression = #list of categories to exclude: error, info, batch, debug, statement, #commit, rollback and result are valid values # (default is info,debug,result,resultset,batch) #excludecategories=info,debug,result,resultset,batch # Execution threshold applies to the standard logging of P6Spy. # While the standard logging logs out every statement # regardless of its execution time, this feature puts a time # condition on that logging. Only statements that have taken # longer than the time specified (in milliseconds) will be # logged. This way it is possible to see only statements that # have exceeded some high water mark. # This time is reloadable. # # executionThreshold=integer time (milliseconds) # (default is 0) #executionThreshold= ################################################################ # P6 OUTAGE SPECIFIC PROPERTIES # ################################################################ # Outage Detection # # This feature detects long-running statements that may be indicative of # a database outage problem. If this feature is turned on, it will log any # statement that surpasses the configurable time boundary during its execution. # When this feature is enabled, no other statements are logged except the long # running statements. The interval property is the boundary time set in seconds. # For example, if this is set to 2, then any statement requiring at least 2 # seconds will be logged. Note that the same statement will continue to be logged # for as long as it executes. So if the interval is set to 2, and the query takes # 11 seconds, it will be logged 5 times (at the 2, 4, 6, 8, 10 second intervals). # # outagedetection=true|false # outagedetectioninterval=integer time (seconds) # # (default is false) #outagedetection=false # (default is 60) #outagedetectioninterval=30
总结：
1.出于安全需要，我改为使用 logger 在 debug 模式下输出。
2.关于使用Slf4j还是Log4j个人推荐使用Slf4j，虽然只是一个简单日志门面，log4j比它强大，但是部署项目时可以方便的变更日志系统
3.很多人可能在配置数据源的时候卡住，p6spy只是代理，其Driver并不能作为数据源的驱动包,故我们使用源Driver即可，在其基础上代理一层即可