pkumarsinha commented on a change in pull request #2083:
URL: https://github.com/apache/hive/pull/2083#discussion_r604942185
##########
File path:
itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosExternalTables.java
##########
@@ -1606,4 +1613,78 @@ public void testDatabaseLevelCopyDisabled() throws
Throwable {
ReplicationTestUtils.assertExternalFileList(Arrays
.asList("a", "b", "c", "newin", "newout"), tuple.dumpLocation,
primary);
}
+
+ @Test
+ public void testDataCopyEndLogAtSource() throws Throwable {
+ testDataCopyEndLog(false);
+ }
+
+ @Test
+ public void testDataCopyEndLogAtTarget() throws Throwable {
+ testDataCopyEndLog(true);
+ }
+
+ public void testDataCopyEndLog(boolean runCopyTasksOnTarget) throws
Throwable {
+ // Get the logger at the root level.
+ Logger logger = LogManager.getLogger("hive.ql.metadata.Hive");
+ LoggerContext ctx = (LoggerContext) LogManager.getContext(false);
+ Configuration config = ctx.getConfiguration();
+ LoggerConfig loggerConfig = config.getLoggerConfig(logger.getName());
+ loggerConfig.setLevel(Level.DEBUG);
+ ctx.updateLoggers();
+ // Create a String Appender to capture log output
+
+ StringAppender appender = StringAppender.createStringAppender("%m");
+ appender.addToLogger(logger.getName(), Level.DEBUG);
+ appender.start();
+
+ List<String> withClause = Arrays.asList("'distcp.options.update'=''",
+ "'" + HiveConf.ConfVars.REPL_RUN_DATA_COPY_TASKS_ON_TARGET.varname
+ + "'='" + runCopyTasksOnTarget + "'");
+
+ // Perform bootstrap dump & load.
+ primary.run("use " + primaryDbName)
+ .run("create external table a (i int)")
+ .run("insert into a values (1),(2),(3),(4)")
+ .run("create external table b (i int)")
+ .run("insert into b values (5),(6),(7),(8)")
+ .dump(primaryDbName, withClause);
+
+ replica.load(replicatedDbName, primaryDbName, withClause)
+ .run("use " + replicatedDbName)
+ .run("show tables like 'a'")
+ .verifyResults(Collections.singletonList("a"))
+ .run("show tables like 'b'")
+ .verifyResults(Collections.singletonList("b"));
+
+ // Check whether the log contains DATA_COPY_END
+ String logStr = appender.getOutput();
+ assertTrue(logStr, logStr.contains("REPL::DATA_COPY_END:"));
+ logStr.indexOf("REPL::DATA_COPY_END:");
+ // Check the log contains DATA_COPY_END after Distcp
+ assertTrue(logStr, logStr.contains("DistCpOptions"));
+ String postLog = logStr.substring(logStr.indexOf("REPL::DATA_COPY_END:"));
+
+ assertFalse(postLog, postLog.contains("DistCpOptions"));
+ appender.reset();
+
+ // Perform incremental dump & load.
+ primary.run("create table c (i int)")
+ .run("insert into c values (10),(11)")
+ .run("insert into a values (5),(6)")
+ .run("insert into b values (9),(10)").dump(primaryDbName, withClause);
+
+ replica.load(replicatedDbName, primaryDbName, withClause)
+ .run("select i From c")
+ .verifyResults(new String[] {"10", "11"});
+
+ // Check whether the log contains DATA_COPY_END
+ logStr = appender.getOutput();
+ assertTrue(logStr, logStr.contains("REPL::DATA_COPY_END:"));
Review comment:
Wouldn't this logStr part be very long one in case of failure. Should we
fix some constant error message like:
assertTrue("The log doesn't contain log line with a word
REPL::DATA_COPY_END", logStr.contains("REPL::DATA_COPY_END:"));
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
