2 # Licensed to the Apache Software Foundation (ASF) under one
3 # or more contributor license agreements. See the NOTICE file
4 # distributed with this work for additional information
5 # regarding copyright ownership. The ASF licenses this file
6 # to you under the Apache License, Version 2.0 (the
7 # "License"); you may not use this file except in compliance
8 # with the License. You may obtain a copy of the License at
10 # http://www.apache.org/licenses/LICENSE-2.0
12 # Unless required by applicable law or agreed to in writing,
13 # software distributed under the License is distributed on an
14 # "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15 # KIND, either express or implied. See the License for the
16 # specific language governing permissions and limitations
21 echo "Usage: ${0} [options] /path/to/component/bin-install /path/to/hadoop/executable /path/to/hadoop/hadoop-yarn-server-tests-tests.jar /path/to/hadoop/hadoop-mapreduce-client-jobclient-tests.jar"
23 echo " --zookeeper-data /path/to/use Where the embedded zookeeper instance should write its data."
24 echo " defaults to 'zk-data' in the working-dir."
25 echo " --working-dir /path/to/use Path for writing configs and logs. must exist."
26 echo " defaults to making a directory via mktemp."
27 echo " --hadoop-client-classpath /path/to/some.jar:/path/to/another.jar classpath for hadoop jars."
28 echo " defaults to 'hadoop classpath'"
29 echo " --hbase-client-install /path/to/unpacked/client/tarball if given we'll look here for hbase client jars instead of the bin-install"
30 echo " --force-data-clean Delete all data in HDFS and ZK prior to starting up hbase"
31 echo " --single-process Run as single process instead of pseudo-distributed"
35 # if no args specified, show usage
41 declare component_install
46 declare distributed
="true"
52 --working-dir) shift; working_dir
=$1; shift;;
53 --force-data-clean) shift; clean
="true";;
54 --zookeeper-data) shift; zk_data_dir
=$1; shift;;
55 --single-process) shift; distributed
="false";;
56 --hadoop-client-classpath) shift; hadoop_jars
="$1"; shift;;
57 --hbase-client-install) shift; hbase_client
="$1"; shift;;
60 *) break;; # terminate while loop
64 # should still have where component checkout is.
68 component_install
="$(cd "$
(dirname "$1")"; pwd)/$(basename "$1")"
69 hadoop_exec
="$(cd "$
(dirname "$2")"; pwd)/$(basename "$2")"
70 yarn_server_tests_test_jar
="$(cd "$
(dirname "$3")"; pwd)/$(basename "$3")"
71 mapred_jobclient_test_jar
="$(cd "$
(dirname "$4")"; pwd)/$(basename "$4")"
73 if [ ! -x "${hadoop_exec}" ]; then
74 echo "hadoop cli does not appear to be executable." >&2
78 if [ ! -d "${component_install}" ]; then
79 echo "Path to HBase binary install should be a directory." >&2
83 if [ ! -f "${yarn_server_tests_test_jar}" ]; then
84 echo "Specified YARN server tests test jar is not a file." >&2
88 if [ ! -f "${mapred_jobclient_test_jar}" ]; then
89 echo "Specified MapReduce jobclient test jar is not a file." >&2
93 if [ -z "${working_dir}" ]; then
94 if ! working_dir
="$(mktemp -d -t hbase-pseudo-dist-test)" ; then
95 echo "Failed to create temporary working directory. Please specify via --working-dir" >&2
100 working_dir
="$(cd "$
(dirname "${working_dir}")"; pwd)/$(basename "${working_dir}")"
101 if [ ! -d "${working_dir}" ]; then
102 echo "passed working directory '${working_dir}' must already exist." >&2
107 if [ -z "${zk_data_dir}" ]; then
108 zk_data_dir
="${working_dir}/zk-data"
109 mkdir
"${zk_data_dir}"
112 zk_data_dir
="$(cd "$
(dirname "${zk_data_dir}")"; pwd)/$(basename "${zk_data_dir}")"
113 if [ ! -d "${zk_data_dir}" ]; then
114 echo "passed directory for unpacking the source tarball '${zk_data_dir}' must already exist."
119 if [ -z "${hbase_client}" ]; then
120 hbase_client
="${component_install}"
122 echo "Using HBase client-side artifact"
124 hbase_client
="$(cd "$
(dirname "${hbase_client}")"; pwd)/$(basename "${hbase_client}")"
125 if [ ! -d "${hbase_client}" ]; then
126 echo "If given hbase client install should be a directory with contents of the client tarball." >&2
131 if [ -n "${hadoop_jars}" ]; then
133 for entry
in $
(echo "${hadoop_jars}" |
tr ':' '\n'); do
134 tmp_jars
=("${tmp_jars[@]}" "$(cd "$(dirname "${entry}")"; pwd)/$(basename "${entry}")")
136 hadoop_jars="$
(IFS
=:; echo "${tmp_jars[*]}")"
140 echo "You
'll find logs and temp files in ${working_dir}"
142 function redirect_and_run {
145 echo "$*" >"${log_base}.err"
146 "$@" >"${log_base}.out" 2>>"${log_base}.err"
151 echo "Hadoop version information:"
152 "${hadoop_exec}" version
153 hadoop_version=$("${hadoop_exec}" version | head -n 1)
154 hadoop_version="${hadoop_version#Hadoop }"
155 if [ "${hadoop_version%.*.*}" -gt 2 ]; then
156 "${hadoop_exec}" envvars
158 echo "JAVA_HOME: ${JAVA_HOME}"
161 # Ensure that if some other Hadoop install happens to be present in the environment we ignore it.
162 HBASE_DISABLE_HADOOP_CLASSPATH_LOOKUP="true"
163 export HBASE_DISABLE_HADOOP_CLASSPATH_LOOKUP
165 if [ -n "${clean}" ]; then
166 echo "Cleaning out ZooKeeper..."
167 rm -rf "${zk_data_dir:?}/*"
170 echo "HBase version information:"
171 "${component_install}/bin/hbase" version 2>/dev/null
172 hbase_version=$("${component_install}/bin/hbase" version | head -n 1 2>/dev/null)
173 hbase_version="${hbase_version#HBase }"
175 if [ ! -s "${hbase_client}/lib/shaded-clients/hbase-shaded-mapreduce-${hbase_version}.jar" ]; then
176 echo "HBase binary install doesn't appear to include a shaded mapreduce artifact.
" >&2
180 if [ ! -s "${hbase_client}/lib
/shaded-clients
/hbase-shaded-client-
${hbase_version}.jar
" ]; then
181 echo "HBase binary
install doesn
't appear to include a shaded client artifact." >&2
185 if [ ! -s "${hbase_client}/lib/shaded-clients/hbase-shaded-client-byo-hadoop-${hbase_version}.jar" ]; then
186 echo "HBase binary install doesn't appear to include a shaded client artifact.
" >&2
190 echo "Writing out configuration
for HBase.
"
191 rm -rf "${working_dir}/hbase-conf
"
192 mkdir "${working_dir}/hbase-conf
"
194 if [ -f "${component_install}/conf
/log4j.properties
" ]; then
195 cp "${component_install}/conf
/log4j.properties
" "${working_dir}/hbase-conf
/log4j.properties
"
197 cat >"${working_dir}/hbase-conf
/log4j.properties
" <<EOF
198 # Define some default values that can be overridden by system properties
199 hbase.root.logger=INFO,console
201 # Define the root logger to the system property "hbase.root.logger
".
202 log4j.rootLogger=${hbase.root.logger}
207 log4j.appender.console=org.apache.log4j.ConsoleAppender
208 log4j.appender.console.target=System.err
209 log4j.appender.console.layout=org.apache.log4j.PatternLayout
210 log4j.appender.console.layout.ConversionPattern=%d{ISO8601} %-5p [%t] %c{2}: %.1000m%n
214 cat >"${working_dir}/hbase-conf
/hbase-site.xml
" <<EOF
215 <?xml version="1.0"?>
216 <?xml-stylesheet type="text
/xsl
" href="configuration.xsl
"?>
220 * Licensed to the Apache Software Foundation (ASF) under one
221 * or more contributor license agreements. See the NOTICE file
222 * distributed with this work for additional information
223 * regarding copyright ownership. The ASF licenses this file
224 * to you under the Apache License, Version 2.0 (the
225 * "License
"); you may not use this file except in compliance
226 * with the License. You may obtain a copy of the License at
228 * http://www.apache.org/licenses/LICENSE-2.0
230 * Unless required by applicable law or agreed to in writing, software
231 * distributed under the License is distributed on an "AS IS
" BASIS,
232 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
233 * See the License for the specific language governing permissions and
234 * limitations under the License.
239 <name>hbase.rootdir</name>
240 <!-- We rely on the defaultFS being set in our hadoop confs -->
241 <value>/hbase</value>
244 <name>hbase.zookeeper.property.dataDir</name>
245 <value>${zk_data_dir}</value>
248 <name>hbase.cluster.distributed</name>
249 <value>${distributed}</value>
254 if [ "true
" = "${distributed}" ]; then
255 cat >"${working_dir}/hbase-conf
/regionservers
" <<EOF
262 echo "Shutting down HBase
"
263 HBASE_CONF_DIR="${working_dir}/hbase-conf
/" "${component_install}/bin
/stop-hbase.sh
"
265 if [ -f "${working_dir}/hadoop.pid
" ]; then
266 echo "Shutdown
: listing HDFS contents
"
267 redirect_and_run "${working_dir}/hadoop_listing_at_end
" \
268 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -ls -R /
270 echo "Shutting down Hadoop
"
271 kill -6 "$
(cat "${working_dir}/hadoop.pid")"
275 trap cleanup EXIT SIGQUIT
277 echo "Starting up Hadoop
"
279 HADOOP_CLASSPATH="${yarn_server_tests_test_jar}" "${hadoop_exec}" jar "${mapred_jobclient_test_jar}" minicluster -writeConfig "${working_dir}/hbase-conf/core-site.xml" -writeDetails "${working_dir}/hadoop_cluster_info.json" >"${working_dir}/hadoop_cluster_command.out" 2>"${working_dir}/hadoop_cluster_command.err
" &
280 echo "$
!" > "${working_dir}/hadoop.pid
"
283 until [ -s "${working_dir}/hbase-conf
/core-site.xml
" ]; do
284 printf '\twaiting for Hadoop to finish starting up.\n'
285 sleep "${sleep_time}"
286 sleep_time="$
((sleep_time
*2))"
289 if [ "${hadoop_version%.*.*}" -gt 2 ]; then
290 echo "Verifying configs
"
291 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" conftest
294 if [ -n "${clean}" ]; then
295 echo "Cleaning out HDFS...
"
296 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -rm -r /hbase
297 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -rm -r example/
298 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -rm -r example-region-listing.data
301 echo "Listing HDFS contents
"
302 redirect_and_run "${working_dir}/hadoop_cluster_smoke
" \
303 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -ls -R /
305 echo "Starting up HBase
"
306 HBASE_CONF_DIR="${working_dir}/hbase-conf
/" "${component_install}/bin
/start-hbase.sh
"
309 until "${component_install}/bin/hbase" --config "${working_dir}/hbase-conf/" shell --noninteractive >"${working_dir}/waiting_hbase_startup.log
" 2>&1 <<EOF
313 printf '\tretry waiting for hbase to come up.\n'
314 sleep "${sleep_time}"
315 sleep_time="$
((sleep_time
*2))"
318 echo "Setting up table
'test:example' with
1,000 regions
"
319 "${hbase_client}/bin/hbase" --config "${working_dir}/hbase-conf/" shell --noninteractive >"${working_dir}/table_create.log
" 2>&1 <<EOF
320 create_namespace 'test'
321 create 'test:example', 'family1', 'family2', {NUMREGIONS => 1000, SPLITALGO => 'UniformSplit'}
324 echo "writing out example TSV to example.tsv
"
325 cat >"${working_dir}/example.tsv
" <<EOF
376 echo "uploading example.tsv to HDFS
"
377 "${hadoop_exec}" --config "${working_dir}/hbase-conf
/" fs -mkdir example
378 "${hadoop_exec}" --config "${working_dir}/hbase-conf/" fs -copyFromLocal "${working_dir}/example.tsv
" "example
/"
380 echo "Importing TSV via shaded client artifact
for HBase
- MapReduce integration.
"
381 # hbase_thirdparty_jars=("${component_install}"/lib/htrace-core4*.jar \
382 # "${component_install}"/lib/slf4j-api-*.jar \
383 # "${component_install}"/lib/commons-logging-*.jar \
384 # "${component_install}"/lib/slf4j-log4j12-*.jar \
385 # "${component_install}"/lib/log4j-1.2.*.jar \
386 # "${working_dir}/hbase-conf
/log4j.properties
")
387 # hbase_dep_classpath=$(IFS=:; echo "${hbase_thirdparty_jars[*]}")
388 hbase_dep_classpath="$
("${hbase_client}/bin/hbase" --config "${working_dir}/hbase-conf/" mapredcp
)"
389 HADOOP_CLASSPATH="${hbase_dep_classpath}" redirect_and_run "${working_dir}/mr-importtsv
" \
390 "${hadoop_exec}" --config "${working_dir}/hbase-conf/" jar "${hbase_client}/lib/shaded-clients/hbase-shaded-mapreduce-${hbase_version}.jar" importtsv -Dimporttsv.columns=HBASE_ROW_KEY,family1:column1,family1:column4,family1:column3 test:example example/ -libjars "${hbase_dep_classpath}"
391 "${hbase_client}/bin/hbase" --config "${working_dir}/hbase-conf/" shell --noninteractive >"${working_dir}/scan_import.out" 2>"${working_dir}/scan_import.err
" <<EOF
395 echo "Verifying row count from import.
"
396 import_rowcount=$(echo 'count "test:example
"' | "${hbase_client}/bin
/hbase
" --config "${working_dir}/hbase-conf
/" shell --noninteractive 2>/dev/null | tail -n 1)
397 if [ ! "${import_rowcount}" -eq 48 ]; then
398 echo "ERROR
: Instead of finding
48 rows
, we found
${import_rowcount}.
"
402 if [ -z "${hadoop_jars}" ]; then
403 echo "Hadoop client jars not given
; getting them from
'hadoop classpath' for the example.
"
404 hadoop_jars=$("${hadoop_exec}" --config "${working_dir}/hbase-conf
/" classpath)
407 echo "Building shaded client example.
"
408 cat >"${working_dir}/HBaseClientReadWriteExample.java
" <<EOF
409 import org.apache.hadoop.conf.Configuration;
410 import org.apache.hadoop.fs.FileSystem;
411 import org.apache.hadoop.fs.FSDataInputStream;
412 import org.apache.hadoop.fs.FSDataOutputStream;
413 import org.apache.hadoop.fs.Path;
414 import org.apache.hadoop.hbase.Cell;
415 import org.apache.hadoop.hbase.CellBuilder;
416 import org.apache.hadoop.hbase.CellBuilderFactory;
417 import org.apache.hadoop.hbase.CellBuilderType;
418 import org.apache.hadoop.hbase.ClusterMetrics;
419 import org.apache.hadoop.hbase.HBaseConfiguration;
420 import org.apache.hadoop.hbase.RegionMetrics;
421 import org.apache.hadoop.hbase.ServerMetrics;
422 import org.apache.hadoop.hbase.TableName;
423 import org.apache.hadoop.hbase.client.Admin;
424 import org.apache.hadoop.hbase.client.Connection;
425 import org.apache.hadoop.hbase.client.ConnectionFactory;
426 import org.apache.hadoop.hbase.client.Put;
427 import org.apache.hadoop.hbase.client.Table;
428 import org.apache.hadoop.hbase.util.Bytes;
430 import java.util.LinkedList;
431 import java.util.List;
434 public class HBaseClientReadWriteExample {
435 private static final byte[] FAMILY_BYTES = Bytes.toBytes("family2
");
437 public static void main(String[] args) throws Exception {
438 Configuration hbase = HBaseConfiguration.create();
439 Configuration hadoop = new Configuration();
440 try (Connection connection = ConnectionFactory.createConnection(hbase)) {
441 System.out.println("Generating list of regions
");
442 final List<String> regions = new LinkedList<>();
443 try (Admin admin = connection.getAdmin()) {
444 final ClusterMetrics cluster = admin.getClusterMetrics();
445 System.out.println(String.format("\tCluster reports version
%s
, ave load
%f
, region count
%d
", cluster.getHBaseVersion(), cluster.getAverageLoad(), cluster.getRegionCount()));
446 for (ServerMetrics server : cluster.getLiveServerMetrics().values()) {
447 for (RegionMetrics region : server.getRegionMetrics().values()) {
448 regions.add(region.getNameAsString());
452 final Path listing = new Path("example-region-listing.data
");
453 System.out.println("Writing list to HDFS
");
454 try (FileSystem fs = FileSystem.newInstance(hadoop)) {
455 final Path path = fs.makeQualified(listing);
456 try (FSDataOutputStream out = fs.create(path)) {
457 out.writeInt(regions.size());
458 for (String region : regions) {
459 out.writeUTF(region);
464 final List<Put> puts = new LinkedList<>();
465 final Put marker = new Put(new byte[] { (byte)0 });
466 System.out.println("Reading list from HDFS
");
467 try (FileSystem fs = FileSystem.newInstance(hadoop)) {
468 final Path path = fs.makeQualified(listing);
469 final CellBuilder builder = CellBuilderFactory.create(CellBuilderType.SHALLOW_COPY);
470 try (FSDataInputStream in = fs.open(path)) {
471 final int count = in.readInt();
472 marker.addColumn(FAMILY_BYTES, Bytes.toBytes("count
"), Bytes.toBytes(count));
473 for(int i = 0; i < count; i++) {
475 final byte[] row = Bytes.toBytes(in.readUTF());
476 final Put put = new Put(row);
478 builder.setFamily(FAMILY_BYTES);
479 builder.setType(Cell.Type.Put);
480 put.add(builder.build());
485 System.out.println("Writing list into HBase table
");
486 try (Table table = connection.getTable(TableName.valueOf("test:example
"))) {
494 redirect_and_run "${working_dir}/hbase-shaded-client-compile
" \
495 javac -cp "${hbase_client}/lib/shaded-clients/hbase-shaded-client-byo-hadoop-${hbase_version}.jar:${hadoop_jars}" "${working_dir}/HBaseClientReadWriteExample.java
"
496 echo "Running shaded client example. It
'll fetch the set of regions, round-trip them to a file in HDFS, then write them one-per-row into the test table."
497 # The order of classpath entries here is important. if we're using non-shaded Hadoop
3 / 2.9.0 jars
, we have to work around YARN-2190.
498 redirect_and_run
"${working_dir}/hbase-shaded-client-example" \
499 java
-cp "${working_dir}/hbase-conf/:${hbase_client}/lib/shaded-clients/hbase-shaded-client-byo-hadoop-${hbase_version}.jar:${hbase_dep_classpath}:${working_dir}:${hadoop_jars}" HBaseClientReadWriteExample
501 echo "Checking on results of example program."
502 "${hadoop_exec}" --config "${working_dir}/hbase-conf/" fs -copyToLocal "example-region-listing.data" "${working_dir}/example-region-listing.data"
504 "${hbase_client}/bin/hbase" --config "${working_dir}/hbase-conf/" shell --noninteractive >"${working_dir}/scan_example.out" 2>"${working_dir}/scan_example.err" <<EOF
508 echo "Verifying row count from example."
509 example_rowcount
=$
(echo 'count "test:example"' |
"${hbase_client}/bin/hbase" --config "${working_dir}/hbase-conf/" shell
--noninteractive 2>/dev
/null |
tail -n 1)
510 if [ "${example_rowcount}" -gt "1050" ]; then
511 echo "Found ${example_rowcount} rows, which is enough to cover 48 for import, 1000 example's use of user table regions, 2 for example's use of meta/root regions, and 1 for example's count record"
513 echo "ERROR: Only found ${example_rowcount} rows."