blob: eeeb98839c432124a717efab44275040626edbc1 [file] [log] [blame]
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html lang="en">
<head>
<title>Source code</title>
<link rel="stylesheet" type="text/css" href="../../../../../../stylesheet.css" title="Style">
</head>
<body>
<div class="sourceContainer">
<pre><span class="sourceLineNo">001</span>/*<a name="line.1"></a>
<span class="sourceLineNo">002</span> * Licensed to the Apache Software Foundation (ASF) under one<a name="line.2"></a>
<span class="sourceLineNo">003</span> * or more contributor license agreements. See the NOTICE file<a name="line.3"></a>
<span class="sourceLineNo">004</span> * distributed with this work for additional information<a name="line.4"></a>
<span class="sourceLineNo">005</span> * regarding copyright ownership. The ASF licenses this file<a name="line.5"></a>
<span class="sourceLineNo">006</span> * to you under the Apache License, Version 2.0 (the<a name="line.6"></a>
<span class="sourceLineNo">007</span> * "License"); you may not use this file except in compliance<a name="line.7"></a>
<span class="sourceLineNo">008</span> * with the License. You may obtain a copy of the License at<a name="line.8"></a>
<span class="sourceLineNo">009</span> *<a name="line.9"></a>
<span class="sourceLineNo">010</span> * http://www.apache.org/licenses/LICENSE-2.0<a name="line.10"></a>
<span class="sourceLineNo">011</span> *<a name="line.11"></a>
<span class="sourceLineNo">012</span> * Unless required by applicable law or agreed to in writing, software<a name="line.12"></a>
<span class="sourceLineNo">013</span> * distributed under the License is distributed on an "AS IS" BASIS,<a name="line.13"></a>
<span class="sourceLineNo">014</span> * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.<a name="line.14"></a>
<span class="sourceLineNo">015</span> * See the License for the specific language governing permissions and<a name="line.15"></a>
<span class="sourceLineNo">016</span> * limitations under the License.<a name="line.16"></a>
<span class="sourceLineNo">017</span> */<a name="line.17"></a>
<span class="sourceLineNo">018</span>package org.apache.hadoop.hbase.tool;<a name="line.18"></a>
<span class="sourceLineNo">019</span><a name="line.19"></a>
<span class="sourceLineNo">020</span>import static org.junit.Assert.assertArrayEquals;<a name="line.20"></a>
<span class="sourceLineNo">021</span>import static org.junit.Assert.assertEquals;<a name="line.21"></a>
<span class="sourceLineNo">022</span>import static org.junit.Assert.assertNotNull;<a name="line.22"></a>
<span class="sourceLineNo">023</span>import static org.junit.Assert.assertNull;<a name="line.23"></a>
<span class="sourceLineNo">024</span>import static org.junit.Assert.assertTrue;<a name="line.24"></a>
<span class="sourceLineNo">025</span>import static org.junit.Assert.fail;<a name="line.25"></a>
<span class="sourceLineNo">026</span>import static org.mockito.ArgumentMatchers.any;<a name="line.26"></a>
<span class="sourceLineNo">027</span>import static org.mockito.ArgumentMatchers.anyBoolean;<a name="line.27"></a>
<span class="sourceLineNo">028</span>import static org.mockito.ArgumentMatchers.anyList;<a name="line.28"></a>
<span class="sourceLineNo">029</span>import static org.mockito.Mockito.doReturn;<a name="line.29"></a>
<span class="sourceLineNo">030</span>import static org.mockito.Mockito.spy;<a name="line.30"></a>
<span class="sourceLineNo">031</span><a name="line.31"></a>
<span class="sourceLineNo">032</span>import java.io.IOException;<a name="line.32"></a>
<span class="sourceLineNo">033</span>import java.nio.ByteBuffer;<a name="line.33"></a>
<span class="sourceLineNo">034</span>import java.util.Deque;<a name="line.34"></a>
<span class="sourceLineNo">035</span>import java.util.List;<a name="line.35"></a>
<span class="sourceLineNo">036</span>import java.util.Map;<a name="line.36"></a>
<span class="sourceLineNo">037</span>import java.util.concurrent.CompletableFuture;<a name="line.37"></a>
<span class="sourceLineNo">038</span>import java.util.concurrent.atomic.AtomicInteger;<a name="line.38"></a>
<span class="sourceLineNo">039</span>import java.util.stream.IntStream;<a name="line.39"></a>
<span class="sourceLineNo">040</span>import org.apache.hadoop.conf.Configuration;<a name="line.40"></a>
<span class="sourceLineNo">041</span>import org.apache.hadoop.fs.FileSystem;<a name="line.41"></a>
<span class="sourceLineNo">042</span>import org.apache.hadoop.fs.Path;<a name="line.42"></a>
<span class="sourceLineNo">043</span>import org.apache.hadoop.hbase.HBaseClassTestRule;<a name="line.43"></a>
<span class="sourceLineNo">044</span>import org.apache.hadoop.hbase.HBaseTestingUtil;<a name="line.44"></a>
<span class="sourceLineNo">045</span>import org.apache.hadoop.hbase.HConstants;<a name="line.45"></a>
<span class="sourceLineNo">046</span>import org.apache.hadoop.hbase.TableExistsException;<a name="line.46"></a>
<span class="sourceLineNo">047</span>import org.apache.hadoop.hbase.TableName;<a name="line.47"></a>
<span class="sourceLineNo">048</span>import org.apache.hadoop.hbase.client.Admin;<a name="line.48"></a>
<span class="sourceLineNo">049</span>import org.apache.hadoop.hbase.client.AsyncClusterConnection;<a name="line.49"></a>
<span class="sourceLineNo">050</span>import org.apache.hadoop.hbase.client.ColumnFamilyDescriptorBuilder;<a name="line.50"></a>
<span class="sourceLineNo">051</span>import org.apache.hadoop.hbase.client.Connection;<a name="line.51"></a>
<span class="sourceLineNo">052</span>import org.apache.hadoop.hbase.client.RegionInfo;<a name="line.52"></a>
<span class="sourceLineNo">053</span>import org.apache.hadoop.hbase.client.Result;<a name="line.53"></a>
<span class="sourceLineNo">054</span>import org.apache.hadoop.hbase.client.ResultScanner;<a name="line.54"></a>
<span class="sourceLineNo">055</span>import org.apache.hadoop.hbase.client.Scan;<a name="line.55"></a>
<span class="sourceLineNo">056</span>import org.apache.hadoop.hbase.client.Table;<a name="line.56"></a>
<span class="sourceLineNo">057</span>import org.apache.hadoop.hbase.client.TableDescriptor;<a name="line.57"></a>
<span class="sourceLineNo">058</span>import org.apache.hadoop.hbase.client.TableDescriptorBuilder;<a name="line.58"></a>
<span class="sourceLineNo">059</span>import org.apache.hadoop.hbase.coprocessor.CoprocessorHost;<a name="line.59"></a>
<span class="sourceLineNo">060</span>import org.apache.hadoop.hbase.regionserver.HRegionServer;<a name="line.60"></a>
<span class="sourceLineNo">061</span>import org.apache.hadoop.hbase.regionserver.TestHRegionServerBulkLoad;<a name="line.61"></a>
<span class="sourceLineNo">062</span>import org.apache.hadoop.hbase.testclassification.LargeTests;<a name="line.62"></a>
<span class="sourceLineNo">063</span>import org.apache.hadoop.hbase.testclassification.MiscTests;<a name="line.63"></a>
<span class="sourceLineNo">064</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.64"></a>
<span class="sourceLineNo">065</span>import org.apache.hadoop.hbase.util.CommonFSUtils;<a name="line.65"></a>
<span class="sourceLineNo">066</span>import org.apache.hadoop.hbase.util.Pair;<a name="line.66"></a>
<span class="sourceLineNo">067</span>import org.junit.AfterClass;<a name="line.67"></a>
<span class="sourceLineNo">068</span>import org.junit.BeforeClass;<a name="line.68"></a>
<span class="sourceLineNo">069</span>import org.junit.ClassRule;<a name="line.69"></a>
<span class="sourceLineNo">070</span>import org.junit.Rule;<a name="line.70"></a>
<span class="sourceLineNo">071</span>import org.junit.Test;<a name="line.71"></a>
<span class="sourceLineNo">072</span>import org.junit.experimental.categories.Category;<a name="line.72"></a>
<span class="sourceLineNo">073</span>import org.junit.rules.TestName;<a name="line.73"></a>
<span class="sourceLineNo">074</span>import org.slf4j.Logger;<a name="line.74"></a>
<span class="sourceLineNo">075</span>import org.slf4j.LoggerFactory;<a name="line.75"></a>
<span class="sourceLineNo">076</span><a name="line.76"></a>
<span class="sourceLineNo">077</span>import org.apache.hbase.thirdparty.com.google.common.collect.Multimap;<a name="line.77"></a>
<span class="sourceLineNo">078</span><a name="line.78"></a>
<span class="sourceLineNo">079</span>import org.apache.hadoop.hbase.shaded.protobuf.ProtobufUtil;<a name="line.79"></a>
<span class="sourceLineNo">080</span><a name="line.80"></a>
<span class="sourceLineNo">081</span>/**<a name="line.81"></a>
<span class="sourceLineNo">082</span> * Test cases for the atomic load error handling of the bulk load functionality.<a name="line.82"></a>
<span class="sourceLineNo">083</span> */<a name="line.83"></a>
<span class="sourceLineNo">084</span>@Category({ MiscTests.class, LargeTests.class })<a name="line.84"></a>
<span class="sourceLineNo">085</span>public class TestBulkLoadHFilesSplitRecovery {<a name="line.85"></a>
<span class="sourceLineNo">086</span><a name="line.86"></a>
<span class="sourceLineNo">087</span> @ClassRule<a name="line.87"></a>
<span class="sourceLineNo">088</span> public static final HBaseClassTestRule CLASS_RULE =<a name="line.88"></a>
<span class="sourceLineNo">089</span> HBaseClassTestRule.forClass(TestBulkLoadHFilesSplitRecovery.class);<a name="line.89"></a>
<span class="sourceLineNo">090</span><a name="line.90"></a>
<span class="sourceLineNo">091</span> private static final Logger LOG = LoggerFactory.getLogger(TestHRegionServerBulkLoad.class);<a name="line.91"></a>
<span class="sourceLineNo">092</span><a name="line.92"></a>
<span class="sourceLineNo">093</span> static HBaseTestingUtil util;<a name="line.93"></a>
<span class="sourceLineNo">094</span> // used by secure subclass<a name="line.94"></a>
<span class="sourceLineNo">095</span> static boolean useSecure = false;<a name="line.95"></a>
<span class="sourceLineNo">096</span><a name="line.96"></a>
<span class="sourceLineNo">097</span> final static int NUM_CFS = 10;<a name="line.97"></a>
<span class="sourceLineNo">098</span> final static byte[] QUAL = Bytes.toBytes("qual");<a name="line.98"></a>
<span class="sourceLineNo">099</span> final static int ROWCOUNT = 100;<a name="line.99"></a>
<span class="sourceLineNo">100</span><a name="line.100"></a>
<span class="sourceLineNo">101</span> private final static byte[][] families = new byte[NUM_CFS][];<a name="line.101"></a>
<span class="sourceLineNo">102</span><a name="line.102"></a>
<span class="sourceLineNo">103</span> @Rule<a name="line.103"></a>
<span class="sourceLineNo">104</span> public TestName name = new TestName();<a name="line.104"></a>
<span class="sourceLineNo">105</span><a name="line.105"></a>
<span class="sourceLineNo">106</span> static {<a name="line.106"></a>
<span class="sourceLineNo">107</span> for (int i = 0; i &lt; NUM_CFS; i++) {<a name="line.107"></a>
<span class="sourceLineNo">108</span> families[i] = Bytes.toBytes(family(i));<a name="line.108"></a>
<span class="sourceLineNo">109</span> }<a name="line.109"></a>
<span class="sourceLineNo">110</span> }<a name="line.110"></a>
<span class="sourceLineNo">111</span><a name="line.111"></a>
<span class="sourceLineNo">112</span> static byte[] rowkey(int i) {<a name="line.112"></a>
<span class="sourceLineNo">113</span> return Bytes.toBytes(String.format("row_%08d", i));<a name="line.113"></a>
<span class="sourceLineNo">114</span> }<a name="line.114"></a>
<span class="sourceLineNo">115</span><a name="line.115"></a>
<span class="sourceLineNo">116</span> static String family(int i) {<a name="line.116"></a>
<span class="sourceLineNo">117</span> return String.format("family_%04d", i);<a name="line.117"></a>
<span class="sourceLineNo">118</span> }<a name="line.118"></a>
<span class="sourceLineNo">119</span><a name="line.119"></a>
<span class="sourceLineNo">120</span> static byte[] value(int i) {<a name="line.120"></a>
<span class="sourceLineNo">121</span> return Bytes.toBytes(String.format("%010d", i));<a name="line.121"></a>
<span class="sourceLineNo">122</span> }<a name="line.122"></a>
<span class="sourceLineNo">123</span><a name="line.123"></a>
<span class="sourceLineNo">124</span> public static void buildHFiles(FileSystem fs, Path dir, int value) throws IOException {<a name="line.124"></a>
<span class="sourceLineNo">125</span> byte[] val = value(value);<a name="line.125"></a>
<span class="sourceLineNo">126</span> for (int i = 0; i &lt; NUM_CFS; i++) {<a name="line.126"></a>
<span class="sourceLineNo">127</span> Path testIn = new Path(dir, family(i));<a name="line.127"></a>
<span class="sourceLineNo">128</span><a name="line.128"></a>
<span class="sourceLineNo">129</span> TestHRegionServerBulkLoad.createHFile(fs, new Path(testIn, "hfile_" + i),<a name="line.129"></a>
<span class="sourceLineNo">130</span> Bytes.toBytes(family(i)), QUAL, val, ROWCOUNT);<a name="line.130"></a>
<span class="sourceLineNo">131</span> }<a name="line.131"></a>
<span class="sourceLineNo">132</span> }<a name="line.132"></a>
<span class="sourceLineNo">133</span><a name="line.133"></a>
<span class="sourceLineNo">134</span> private TableDescriptor createTableDesc(TableName name, int cfs) {<a name="line.134"></a>
<span class="sourceLineNo">135</span> TableDescriptorBuilder builder = TableDescriptorBuilder.newBuilder(name);<a name="line.135"></a>
<span class="sourceLineNo">136</span> IntStream.range(0, cfs).mapToObj(i -&gt; ColumnFamilyDescriptorBuilder.of(family(i)))<a name="line.136"></a>
<span class="sourceLineNo">137</span> .forEachOrdered(builder::setColumnFamily);<a name="line.137"></a>
<span class="sourceLineNo">138</span> return builder.build();<a name="line.138"></a>
<span class="sourceLineNo">139</span> }<a name="line.139"></a>
<span class="sourceLineNo">140</span><a name="line.140"></a>
<span class="sourceLineNo">141</span> /**<a name="line.141"></a>
<span class="sourceLineNo">142</span> * Creates a table with given table name and specified number of column families if the table does<a name="line.142"></a>
<span class="sourceLineNo">143</span> * not already exist.<a name="line.143"></a>
<span class="sourceLineNo">144</span> */<a name="line.144"></a>
<span class="sourceLineNo">145</span> private void setupTable(final Connection connection, TableName table, int cfs)<a name="line.145"></a>
<span class="sourceLineNo">146</span> throws IOException {<a name="line.146"></a>
<span class="sourceLineNo">147</span> try {<a name="line.147"></a>
<span class="sourceLineNo">148</span> LOG.info("Creating table " + table);<a name="line.148"></a>
<span class="sourceLineNo">149</span> try (Admin admin = connection.getAdmin()) {<a name="line.149"></a>
<span class="sourceLineNo">150</span> admin.createTable(createTableDesc(table, cfs));<a name="line.150"></a>
<span class="sourceLineNo">151</span> }<a name="line.151"></a>
<span class="sourceLineNo">152</span> } catch (TableExistsException tee) {<a name="line.152"></a>
<span class="sourceLineNo">153</span> LOG.info("Table " + table + " already exists");<a name="line.153"></a>
<span class="sourceLineNo">154</span> }<a name="line.154"></a>
<span class="sourceLineNo">155</span> }<a name="line.155"></a>
<span class="sourceLineNo">156</span><a name="line.156"></a>
<span class="sourceLineNo">157</span> /**<a name="line.157"></a>
<span class="sourceLineNo">158</span> * Creates a table with given table name,specified number of column families&lt;br&gt;<a name="line.158"></a>
<span class="sourceLineNo">159</span> * and splitkeys if the table does not already exist. nnn<a name="line.159"></a>
<span class="sourceLineNo">160</span> */<a name="line.160"></a>
<span class="sourceLineNo">161</span> private void setupTableWithSplitkeys(TableName table, int cfs, byte[][] SPLIT_KEYS)<a name="line.161"></a>
<span class="sourceLineNo">162</span> throws IOException {<a name="line.162"></a>
<span class="sourceLineNo">163</span> try {<a name="line.163"></a>
<span class="sourceLineNo">164</span> LOG.info("Creating table " + table);<a name="line.164"></a>
<span class="sourceLineNo">165</span> util.createTable(createTableDesc(table, cfs), SPLIT_KEYS);<a name="line.165"></a>
<span class="sourceLineNo">166</span> } catch (TableExistsException tee) {<a name="line.166"></a>
<span class="sourceLineNo">167</span> LOG.info("Table " + table + " already exists");<a name="line.167"></a>
<span class="sourceLineNo">168</span> }<a name="line.168"></a>
<span class="sourceLineNo">169</span> }<a name="line.169"></a>
<span class="sourceLineNo">170</span><a name="line.170"></a>
<span class="sourceLineNo">171</span> private Path buildBulkFiles(TableName table, int value) throws Exception {<a name="line.171"></a>
<span class="sourceLineNo">172</span> Path dir = util.getDataTestDirOnTestFS(table.getNameAsString());<a name="line.172"></a>
<span class="sourceLineNo">173</span> Path bulk1 = new Path(dir, table.getNameAsString() + value);<a name="line.173"></a>
<span class="sourceLineNo">174</span> FileSystem fs = util.getTestFileSystem();<a name="line.174"></a>
<span class="sourceLineNo">175</span> buildHFiles(fs, bulk1, value);<a name="line.175"></a>
<span class="sourceLineNo">176</span> return bulk1;<a name="line.176"></a>
<span class="sourceLineNo">177</span> }<a name="line.177"></a>
<span class="sourceLineNo">178</span><a name="line.178"></a>
<span class="sourceLineNo">179</span> /**<a name="line.179"></a>
<span class="sourceLineNo">180</span> * Populate table with known values.<a name="line.180"></a>
<span class="sourceLineNo">181</span> */<a name="line.181"></a>
<span class="sourceLineNo">182</span> private void populateTable(final Connection connection, TableName table, int value)<a name="line.182"></a>
<span class="sourceLineNo">183</span> throws Exception {<a name="line.183"></a>
<span class="sourceLineNo">184</span> // create HFiles for different column families<a name="line.184"></a>
<span class="sourceLineNo">185</span> Path dir = buildBulkFiles(table, value);<a name="line.185"></a>
<span class="sourceLineNo">186</span> BulkLoadHFiles.create(util.getConfiguration()).bulkLoad(table, dir);<a name="line.186"></a>
<span class="sourceLineNo">187</span> }<a name="line.187"></a>
<span class="sourceLineNo">188</span><a name="line.188"></a>
<span class="sourceLineNo">189</span> /**<a name="line.189"></a>
<span class="sourceLineNo">190</span> * Split the known table in half. (this is hard coded for this test suite)<a name="line.190"></a>
<span class="sourceLineNo">191</span> */<a name="line.191"></a>
<span class="sourceLineNo">192</span> private void forceSplit(TableName table) {<a name="line.192"></a>
<span class="sourceLineNo">193</span> try {<a name="line.193"></a>
<span class="sourceLineNo">194</span> // need to call regions server to by synchronous but isn't visible.<a name="line.194"></a>
<span class="sourceLineNo">195</span> HRegionServer hrs = util.getRSForFirstRegionInTable(table);<a name="line.195"></a>
<span class="sourceLineNo">196</span><a name="line.196"></a>
<span class="sourceLineNo">197</span> for (RegionInfo hri : ProtobufUtil.getOnlineRegions(hrs.getRSRpcServices())) {<a name="line.197"></a>
<span class="sourceLineNo">198</span> if (hri.getTable().equals(table)) {<a name="line.198"></a>
<span class="sourceLineNo">199</span> util.getAdmin().splitRegionAsync(hri.getRegionName(), rowkey(ROWCOUNT / 2));<a name="line.199"></a>
<span class="sourceLineNo">200</span> // ProtobufUtil.split(null, hrs.getRSRpcServices(), hri, rowkey(ROWCOUNT / 2));<a name="line.200"></a>
<span class="sourceLineNo">201</span> }<a name="line.201"></a>
<span class="sourceLineNo">202</span> }<a name="line.202"></a>
<span class="sourceLineNo">203</span><a name="line.203"></a>
<span class="sourceLineNo">204</span> // verify that split completed.<a name="line.204"></a>
<span class="sourceLineNo">205</span> int regions;<a name="line.205"></a>
<span class="sourceLineNo">206</span> do {<a name="line.206"></a>
<span class="sourceLineNo">207</span> regions = 0;<a name="line.207"></a>
<span class="sourceLineNo">208</span> for (RegionInfo hri : ProtobufUtil.getOnlineRegions(hrs.getRSRpcServices())) {<a name="line.208"></a>
<span class="sourceLineNo">209</span> if (hri.getTable().equals(table)) {<a name="line.209"></a>
<span class="sourceLineNo">210</span> regions++;<a name="line.210"></a>
<span class="sourceLineNo">211</span> }<a name="line.211"></a>
<span class="sourceLineNo">212</span> }<a name="line.212"></a>
<span class="sourceLineNo">213</span> if (regions != 2) {<a name="line.213"></a>
<span class="sourceLineNo">214</span> LOG.info("Taking some time to complete split...");<a name="line.214"></a>
<span class="sourceLineNo">215</span> Thread.sleep(250);<a name="line.215"></a>
<span class="sourceLineNo">216</span> }<a name="line.216"></a>
<span class="sourceLineNo">217</span> } while (regions != 2);<a name="line.217"></a>
<span class="sourceLineNo">218</span> } catch (IOException e) {<a name="line.218"></a>
<span class="sourceLineNo">219</span> e.printStackTrace();<a name="line.219"></a>
<span class="sourceLineNo">220</span> } catch (InterruptedException e) {<a name="line.220"></a>
<span class="sourceLineNo">221</span> e.printStackTrace();<a name="line.221"></a>
<span class="sourceLineNo">222</span> }<a name="line.222"></a>
<span class="sourceLineNo">223</span> }<a name="line.223"></a>
<span class="sourceLineNo">224</span><a name="line.224"></a>
<span class="sourceLineNo">225</span> @BeforeClass<a name="line.225"></a>
<span class="sourceLineNo">226</span> public static void setupCluster() throws Exception {<a name="line.226"></a>
<span class="sourceLineNo">227</span> util = new HBaseTestingUtil();<a name="line.227"></a>
<span class="sourceLineNo">228</span> util.getConfiguration().set(CoprocessorHost.REGION_COPROCESSOR_CONF_KEY, "");<a name="line.228"></a>
<span class="sourceLineNo">229</span> util.startMiniCluster(1);<a name="line.229"></a>
<span class="sourceLineNo">230</span> }<a name="line.230"></a>
<span class="sourceLineNo">231</span><a name="line.231"></a>
<span class="sourceLineNo">232</span> @AfterClass<a name="line.232"></a>
<span class="sourceLineNo">233</span> public static void teardownCluster() throws Exception {<a name="line.233"></a>
<span class="sourceLineNo">234</span> util.shutdownMiniCluster();<a name="line.234"></a>
<span class="sourceLineNo">235</span> }<a name="line.235"></a>
<span class="sourceLineNo">236</span><a name="line.236"></a>
<span class="sourceLineNo">237</span> /**<a name="line.237"></a>
<span class="sourceLineNo">238</span> * Checks that all columns have the expected value and that there is the expected number of rows.<a name="line.238"></a>
<span class="sourceLineNo">239</span> * n<a name="line.239"></a>
<span class="sourceLineNo">240</span> */<a name="line.240"></a>
<span class="sourceLineNo">241</span> void assertExpectedTable(TableName table, int count, int value) throws IOException {<a name="line.241"></a>
<span class="sourceLineNo">242</span> TableDescriptor htd = util.getAdmin().getDescriptor(table);<a name="line.242"></a>
<span class="sourceLineNo">243</span> assertNotNull(htd);<a name="line.243"></a>
<span class="sourceLineNo">244</span> try (Table t = util.getConnection().getTable(table);<a name="line.244"></a>
<span class="sourceLineNo">245</span> ResultScanner sr = t.getScanner(new Scan())) {<a name="line.245"></a>
<span class="sourceLineNo">246</span> int i = 0;<a name="line.246"></a>
<span class="sourceLineNo">247</span> for (Result r; (r = sr.next()) != null;) {<a name="line.247"></a>
<span class="sourceLineNo">248</span> r.getNoVersionMap().values().stream().flatMap(m -&gt; m.values().stream())<a name="line.248"></a>
<span class="sourceLineNo">249</span> .forEach(v -&gt; assertArrayEquals(value(value), v));<a name="line.249"></a>
<span class="sourceLineNo">250</span> i++;<a name="line.250"></a>
<span class="sourceLineNo">251</span> }<a name="line.251"></a>
<span class="sourceLineNo">252</span> assertEquals(count, i);<a name="line.252"></a>
<span class="sourceLineNo">253</span> } catch (IOException e) {<a name="line.253"></a>
<span class="sourceLineNo">254</span> fail("Failed due to exception");<a name="line.254"></a>
<span class="sourceLineNo">255</span> }<a name="line.255"></a>
<span class="sourceLineNo">256</span> }<a name="line.256"></a>
<span class="sourceLineNo">257</span><a name="line.257"></a>
<span class="sourceLineNo">258</span> private static &lt;T&gt; CompletableFuture&lt;T&gt; failedFuture(Throwable error) {<a name="line.258"></a>
<span class="sourceLineNo">259</span> CompletableFuture&lt;T&gt; future = new CompletableFuture&lt;&gt;();<a name="line.259"></a>
<span class="sourceLineNo">260</span> future.completeExceptionally(error);<a name="line.260"></a>
<span class="sourceLineNo">261</span> return future;<a name="line.261"></a>
<span class="sourceLineNo">262</span> }<a name="line.262"></a>
<span class="sourceLineNo">263</span><a name="line.263"></a>
<span class="sourceLineNo">264</span> private static AsyncClusterConnection mockAndInjectError(AsyncClusterConnection conn) {<a name="line.264"></a>
<span class="sourceLineNo">265</span> AsyncClusterConnection errConn = spy(conn);<a name="line.265"></a>
<span class="sourceLineNo">266</span> doReturn(failedFuture(new IOException("injecting bulk load error"))).when(errConn).bulkLoad(<a name="line.266"></a>
<span class="sourceLineNo">267</span> any(), anyList(), any(), anyBoolean(), any(), any(), anyBoolean(), anyList(), anyBoolean());<a name="line.267"></a>
<span class="sourceLineNo">268</span> return errConn;<a name="line.268"></a>
<span class="sourceLineNo">269</span> }<a name="line.269"></a>
<span class="sourceLineNo">270</span><a name="line.270"></a>
<span class="sourceLineNo">271</span> /**<a name="line.271"></a>
<span class="sourceLineNo">272</span> * Test that shows that exception thrown from the RS side will result in an exception on the<a name="line.272"></a>
<span class="sourceLineNo">273</span> * LIHFile client.<a name="line.273"></a>
<span class="sourceLineNo">274</span> */<a name="line.274"></a>
<span class="sourceLineNo">275</span> @Test(expected = IOException.class)<a name="line.275"></a>
<span class="sourceLineNo">276</span> public void testBulkLoadPhaseFailure() throws Exception {<a name="line.276"></a>
<span class="sourceLineNo">277</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.277"></a>
<span class="sourceLineNo">278</span> final AtomicInteger attemptedCalls = new AtomicInteger();<a name="line.278"></a>
<span class="sourceLineNo">279</span> Configuration conf = new Configuration(util.getConfiguration());<a name="line.279"></a>
<span class="sourceLineNo">280</span> conf.setInt(HConstants.HBASE_CLIENT_RETRIES_NUMBER, 2);<a name="line.280"></a>
<span class="sourceLineNo">281</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(conf) {<a name="line.281"></a>
<span class="sourceLineNo">282</span><a name="line.282"></a>
<span class="sourceLineNo">283</span> @Override<a name="line.283"></a>
<span class="sourceLineNo">284</span> protected void bulkLoadPhase(AsyncClusterConnection conn, TableName tableName,<a name="line.284"></a>
<span class="sourceLineNo">285</span> Deque&lt;LoadQueueItem&gt; queue, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups,<a name="line.285"></a>
<span class="sourceLineNo">286</span> boolean copyFiles, Map&lt;LoadQueueItem, ByteBuffer&gt; item2RegionMap) throws IOException {<a name="line.286"></a>
<span class="sourceLineNo">287</span> AsyncClusterConnection c =<a name="line.287"></a>
<span class="sourceLineNo">288</span> attemptedCalls.incrementAndGet() == 1 ? mockAndInjectError(conn) : conn;<a name="line.288"></a>
<span class="sourceLineNo">289</span> super.bulkLoadPhase(c, tableName, queue, regionGroups, copyFiles, item2RegionMap);<a name="line.289"></a>
<span class="sourceLineNo">290</span> }<a name="line.290"></a>
<span class="sourceLineNo">291</span> };<a name="line.291"></a>
<span class="sourceLineNo">292</span> Path dir = buildBulkFiles(table, 1);<a name="line.292"></a>
<span class="sourceLineNo">293</span> loader.bulkLoad(table, dir);<a name="line.293"></a>
<span class="sourceLineNo">294</span> }<a name="line.294"></a>
<span class="sourceLineNo">295</span><a name="line.295"></a>
<span class="sourceLineNo">296</span> /**<a name="line.296"></a>
<span class="sourceLineNo">297</span> * Test that shows that exception thrown from the RS side will result in the expected number of<a name="line.297"></a>
<span class="sourceLineNo">298</span> * retries set by ${@link HConstants#HBASE_CLIENT_RETRIES_NUMBER} when<a name="line.298"></a>
<span class="sourceLineNo">299</span> * ${@link BulkLoadHFiles#RETRY_ON_IO_EXCEPTION} is set<a name="line.299"></a>
<span class="sourceLineNo">300</span> */<a name="line.300"></a>
<span class="sourceLineNo">301</span> @Test<a name="line.301"></a>
<span class="sourceLineNo">302</span> public void testRetryOnIOException() throws Exception {<a name="line.302"></a>
<span class="sourceLineNo">303</span> TableName table = TableName.valueOf(name.getMethodName());<a name="line.303"></a>
<span class="sourceLineNo">304</span> AtomicInteger calls = new AtomicInteger(0);<a name="line.304"></a>
<span class="sourceLineNo">305</span> setupTable(util.getConnection(), table, 10);<a name="line.305"></a>
<span class="sourceLineNo">306</span> Configuration conf = new Configuration(util.getConfiguration());<a name="line.306"></a>
<span class="sourceLineNo">307</span> conf.setInt(HConstants.HBASE_CLIENT_RETRIES_NUMBER, 2);<a name="line.307"></a>
<span class="sourceLineNo">308</span> conf.setBoolean(BulkLoadHFiles.RETRY_ON_IO_EXCEPTION, true);<a name="line.308"></a>
<span class="sourceLineNo">309</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(conf) {<a name="line.309"></a>
<span class="sourceLineNo">310</span><a name="line.310"></a>
<span class="sourceLineNo">311</span> @Override<a name="line.311"></a>
<span class="sourceLineNo">312</span> protected void bulkLoadPhase(AsyncClusterConnection conn, TableName tableName,<a name="line.312"></a>
<span class="sourceLineNo">313</span> Deque&lt;LoadQueueItem&gt; queue, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups,<a name="line.313"></a>
<span class="sourceLineNo">314</span> boolean copyFiles, Map&lt;LoadQueueItem, ByteBuffer&gt; item2RegionMap) throws IOException {<a name="line.314"></a>
<span class="sourceLineNo">315</span> if (<a name="line.315"></a>
<span class="sourceLineNo">316</span> calls.get() &lt; conf.getInt(HConstants.HBASE_CLIENT_RETRIES_NUMBER,<a name="line.316"></a>
<span class="sourceLineNo">317</span> HConstants.DEFAULT_HBASE_CLIENT_RETRIES_NUMBER)<a name="line.317"></a>
<span class="sourceLineNo">318</span> ) {<a name="line.318"></a>
<span class="sourceLineNo">319</span> calls.incrementAndGet();<a name="line.319"></a>
<span class="sourceLineNo">320</span> super.bulkLoadPhase(mockAndInjectError(conn), tableName, queue, regionGroups, copyFiles,<a name="line.320"></a>
<span class="sourceLineNo">321</span> item2RegionMap);<a name="line.321"></a>
<span class="sourceLineNo">322</span> } else {<a name="line.322"></a>
<span class="sourceLineNo">323</span> super.bulkLoadPhase(conn, tableName, queue, regionGroups, copyFiles, item2RegionMap);<a name="line.323"></a>
<span class="sourceLineNo">324</span> }<a name="line.324"></a>
<span class="sourceLineNo">325</span> }<a name="line.325"></a>
<span class="sourceLineNo">326</span> };<a name="line.326"></a>
<span class="sourceLineNo">327</span> Path dir = buildBulkFiles(table, 1);<a name="line.327"></a>
<span class="sourceLineNo">328</span> loader.bulkLoad(table, dir);<a name="line.328"></a>
<span class="sourceLineNo">329</span> assertEquals(calls.get(), 2);<a name="line.329"></a>
<span class="sourceLineNo">330</span> }<a name="line.330"></a>
<span class="sourceLineNo">331</span><a name="line.331"></a>
<span class="sourceLineNo">332</span> /**<a name="line.332"></a>
<span class="sourceLineNo">333</span> * This test exercises the path where there is a split after initial validation but before the<a name="line.333"></a>
<span class="sourceLineNo">334</span> * atomic bulk load call. We cannot use presplitting to test this path, so we actually inject a<a name="line.334"></a>
<span class="sourceLineNo">335</span> * split just before the atomic region load.<a name="line.335"></a>
<span class="sourceLineNo">336</span> */<a name="line.336"></a>
<span class="sourceLineNo">337</span> @Test<a name="line.337"></a>
<span class="sourceLineNo">338</span> public void testSplitWhileBulkLoadPhase() throws Exception {<a name="line.338"></a>
<span class="sourceLineNo">339</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.339"></a>
<span class="sourceLineNo">340</span> setupTable(util.getConnection(), table, 10);<a name="line.340"></a>
<span class="sourceLineNo">341</span> populateTable(util.getConnection(), table, 1);<a name="line.341"></a>
<span class="sourceLineNo">342</span> assertExpectedTable(table, ROWCOUNT, 1);<a name="line.342"></a>
<span class="sourceLineNo">343</span><a name="line.343"></a>
<span class="sourceLineNo">344</span> // Now let's cause trouble. This will occur after checks and cause bulk<a name="line.344"></a>
<span class="sourceLineNo">345</span> // files to fail when attempt to atomically import. This is recoverable.<a name="line.345"></a>
<span class="sourceLineNo">346</span> final AtomicInteger attemptedCalls = new AtomicInteger();<a name="line.346"></a>
<span class="sourceLineNo">347</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(util.getConfiguration()) {<a name="line.347"></a>
<span class="sourceLineNo">348</span><a name="line.348"></a>
<span class="sourceLineNo">349</span> @Override<a name="line.349"></a>
<span class="sourceLineNo">350</span> protected void bulkLoadPhase(AsyncClusterConnection conn, TableName tableName,<a name="line.350"></a>
<span class="sourceLineNo">351</span> Deque&lt;LoadQueueItem&gt; queue, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups,<a name="line.351"></a>
<span class="sourceLineNo">352</span> boolean copyFiles, Map&lt;LoadQueueItem, ByteBuffer&gt; item2RegionMap) throws IOException {<a name="line.352"></a>
<span class="sourceLineNo">353</span> int i = attemptedCalls.incrementAndGet();<a name="line.353"></a>
<span class="sourceLineNo">354</span> if (i == 1) {<a name="line.354"></a>
<span class="sourceLineNo">355</span> // On first attempt force a split.<a name="line.355"></a>
<span class="sourceLineNo">356</span> forceSplit(table);<a name="line.356"></a>
<span class="sourceLineNo">357</span> }<a name="line.357"></a>
<span class="sourceLineNo">358</span> super.bulkLoadPhase(conn, tableName, queue, regionGroups, copyFiles, item2RegionMap);<a name="line.358"></a>
<span class="sourceLineNo">359</span> }<a name="line.359"></a>
<span class="sourceLineNo">360</span> };<a name="line.360"></a>
<span class="sourceLineNo">361</span><a name="line.361"></a>
<span class="sourceLineNo">362</span> // create HFiles for different column families<a name="line.362"></a>
<span class="sourceLineNo">363</span> Path dir = buildBulkFiles(table, 2);<a name="line.363"></a>
<span class="sourceLineNo">364</span> loader.bulkLoad(table, dir);<a name="line.364"></a>
<span class="sourceLineNo">365</span><a name="line.365"></a>
<span class="sourceLineNo">366</span> // check that data was loaded<a name="line.366"></a>
<span class="sourceLineNo">367</span> // The three expected attempts are 1) failure because need to split, 2)<a name="line.367"></a>
<span class="sourceLineNo">368</span> // load of split top 3) load of split bottom<a name="line.368"></a>
<span class="sourceLineNo">369</span> assertEquals(3, attemptedCalls.get());<a name="line.369"></a>
<span class="sourceLineNo">370</span> assertExpectedTable(table, ROWCOUNT, 2);<a name="line.370"></a>
<span class="sourceLineNo">371</span> }<a name="line.371"></a>
<span class="sourceLineNo">372</span><a name="line.372"></a>
<span class="sourceLineNo">373</span> /**<a name="line.373"></a>
<span class="sourceLineNo">374</span> * This test splits a table and attempts to bulk load. The bulk import files should be split<a name="line.374"></a>
<span class="sourceLineNo">375</span> * before atomically importing.<a name="line.375"></a>
<span class="sourceLineNo">376</span> */<a name="line.376"></a>
<span class="sourceLineNo">377</span> @Test<a name="line.377"></a>
<span class="sourceLineNo">378</span> public void testGroupOrSplitPresplit() throws Exception {<a name="line.378"></a>
<span class="sourceLineNo">379</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.379"></a>
<span class="sourceLineNo">380</span> setupTable(util.getConnection(), table, 10);<a name="line.380"></a>
<span class="sourceLineNo">381</span> populateTable(util.getConnection(), table, 1);<a name="line.381"></a>
<span class="sourceLineNo">382</span> assertExpectedTable(util.getConnection(), table, ROWCOUNT, 1);<a name="line.382"></a>
<span class="sourceLineNo">383</span> forceSplit(table);<a name="line.383"></a>
<span class="sourceLineNo">384</span><a name="line.384"></a>
<span class="sourceLineNo">385</span> final AtomicInteger countedLqis = new AtomicInteger();<a name="line.385"></a>
<span class="sourceLineNo">386</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(util.getConfiguration()) {<a name="line.386"></a>
<span class="sourceLineNo">387</span><a name="line.387"></a>
<span class="sourceLineNo">388</span> @Override<a name="line.388"></a>
<span class="sourceLineNo">389</span> protected Pair&lt;List&lt;LoadQueueItem&gt;, String&gt; groupOrSplit(AsyncClusterConnection conn,<a name="line.389"></a>
<span class="sourceLineNo">390</span> TableName tableName, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups, LoadQueueItem item,<a name="line.390"></a>
<span class="sourceLineNo">391</span> List&lt;Pair&lt;byte[], byte[]&gt;&gt; startEndKeys) throws IOException {<a name="line.391"></a>
<span class="sourceLineNo">392</span> Pair&lt;List&lt;LoadQueueItem&gt;, String&gt; lqis =<a name="line.392"></a>
<span class="sourceLineNo">393</span> super.groupOrSplit(conn, tableName, regionGroups, item, startEndKeys);<a name="line.393"></a>
<span class="sourceLineNo">394</span> if (lqis != null &amp;&amp; lqis.getFirst() != null) {<a name="line.394"></a>
<span class="sourceLineNo">395</span> countedLqis.addAndGet(lqis.getFirst().size());<a name="line.395"></a>
<span class="sourceLineNo">396</span> }<a name="line.396"></a>
<span class="sourceLineNo">397</span> return lqis;<a name="line.397"></a>
<span class="sourceLineNo">398</span> }<a name="line.398"></a>
<span class="sourceLineNo">399</span> };<a name="line.399"></a>
<span class="sourceLineNo">400</span><a name="line.400"></a>
<span class="sourceLineNo">401</span> // create HFiles for different column families<a name="line.401"></a>
<span class="sourceLineNo">402</span> Path dir = buildBulkFiles(table, 2);<a name="line.402"></a>
<span class="sourceLineNo">403</span> loader.bulkLoad(table, dir);<a name="line.403"></a>
<span class="sourceLineNo">404</span> assertExpectedTable(util.getConnection(), table, ROWCOUNT, 2);<a name="line.404"></a>
<span class="sourceLineNo">405</span> assertEquals(20, countedLqis.get());<a name="line.405"></a>
<span class="sourceLineNo">406</span> }<a name="line.406"></a>
<span class="sourceLineNo">407</span><a name="line.407"></a>
<span class="sourceLineNo">408</span> @Test<a name="line.408"></a>
<span class="sourceLineNo">409</span> public void testCorrectSplitPoint() throws Exception {<a name="line.409"></a>
<span class="sourceLineNo">410</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.410"></a>
<span class="sourceLineNo">411</span> byte[][] SPLIT_KEYS = new byte[][] { Bytes.toBytes("row_00000010"),<a name="line.411"></a>
<span class="sourceLineNo">412</span> Bytes.toBytes("row_00000020"), Bytes.toBytes("row_00000030"), Bytes.toBytes("row_00000040"),<a name="line.412"></a>
<span class="sourceLineNo">413</span> Bytes.toBytes("row_00000050"), Bytes.toBytes("row_00000060"), Bytes.toBytes("row_00000070") };<a name="line.413"></a>
<span class="sourceLineNo">414</span> setupTableWithSplitkeys(table, NUM_CFS, SPLIT_KEYS);<a name="line.414"></a>
<span class="sourceLineNo">415</span><a name="line.415"></a>
<span class="sourceLineNo">416</span> final AtomicInteger bulkloadRpcTimes = new AtomicInteger();<a name="line.416"></a>
<span class="sourceLineNo">417</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(util.getConfiguration()) {<a name="line.417"></a>
<span class="sourceLineNo">418</span><a name="line.418"></a>
<span class="sourceLineNo">419</span> @Override<a name="line.419"></a>
<span class="sourceLineNo">420</span> protected void bulkLoadPhase(AsyncClusterConnection conn, TableName tableName,<a name="line.420"></a>
<span class="sourceLineNo">421</span> Deque&lt;LoadQueueItem&gt; queue, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups,<a name="line.421"></a>
<span class="sourceLineNo">422</span> boolean copyFiles, Map&lt;LoadQueueItem, ByteBuffer&gt; item2RegionMap) throws IOException {<a name="line.422"></a>
<span class="sourceLineNo">423</span> bulkloadRpcTimes.addAndGet(1);<a name="line.423"></a>
<span class="sourceLineNo">424</span> super.bulkLoadPhase(conn, tableName, queue, regionGroups, copyFiles, item2RegionMap);<a name="line.424"></a>
<span class="sourceLineNo">425</span> }<a name="line.425"></a>
<span class="sourceLineNo">426</span> };<a name="line.426"></a>
<span class="sourceLineNo">427</span><a name="line.427"></a>
<span class="sourceLineNo">428</span> Path dir = buildBulkFiles(table, 1);<a name="line.428"></a>
<span class="sourceLineNo">429</span> loader.bulkLoad(table, dir);<a name="line.429"></a>
<span class="sourceLineNo">430</span> // before HBASE-25281 we need invoke bulkload rpc 8 times<a name="line.430"></a>
<span class="sourceLineNo">431</span> assertEquals(4, bulkloadRpcTimes.get());<a name="line.431"></a>
<span class="sourceLineNo">432</span> }<a name="line.432"></a>
<span class="sourceLineNo">433</span><a name="line.433"></a>
<span class="sourceLineNo">434</span> /**<a name="line.434"></a>
<span class="sourceLineNo">435</span> * This test creates a table with many small regions. The bulk load files would be splitted<a name="line.435"></a>
<span class="sourceLineNo">436</span> * multiple times before all of them can be loaded successfully.<a name="line.436"></a>
<span class="sourceLineNo">437</span> */<a name="line.437"></a>
<span class="sourceLineNo">438</span> @Test<a name="line.438"></a>
<span class="sourceLineNo">439</span> public void testSplitTmpFileCleanUp() throws Exception {<a name="line.439"></a>
<span class="sourceLineNo">440</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.440"></a>
<span class="sourceLineNo">441</span> byte[][] SPLIT_KEYS = new byte[][] { Bytes.toBytes("row_00000010"),<a name="line.441"></a>
<span class="sourceLineNo">442</span> Bytes.toBytes("row_00000020"), Bytes.toBytes("row_00000030"), Bytes.toBytes("row_00000040"),<a name="line.442"></a>
<span class="sourceLineNo">443</span> Bytes.toBytes("row_00000050") };<a name="line.443"></a>
<span class="sourceLineNo">444</span> setupTableWithSplitkeys(table, 10, SPLIT_KEYS);<a name="line.444"></a>
<span class="sourceLineNo">445</span><a name="line.445"></a>
<span class="sourceLineNo">446</span> BulkLoadHFiles loader = BulkLoadHFiles.create(util.getConfiguration());<a name="line.446"></a>
<span class="sourceLineNo">447</span><a name="line.447"></a>
<span class="sourceLineNo">448</span> // create HFiles<a name="line.448"></a>
<span class="sourceLineNo">449</span> Path dir = buildBulkFiles(table, 2);<a name="line.449"></a>
<span class="sourceLineNo">450</span> loader.bulkLoad(table, dir);<a name="line.450"></a>
<span class="sourceLineNo">451</span> // family path<a name="line.451"></a>
<span class="sourceLineNo">452</span> Path tmpPath = new Path(dir, family(0));<a name="line.452"></a>
<span class="sourceLineNo">453</span> // TMP_DIR under family path<a name="line.453"></a>
<span class="sourceLineNo">454</span> tmpPath = new Path(tmpPath, BulkLoadHFilesTool.TMP_DIR);<a name="line.454"></a>
<span class="sourceLineNo">455</span> FileSystem fs = dir.getFileSystem(util.getConfiguration());<a name="line.455"></a>
<span class="sourceLineNo">456</span> // HFiles have been splitted, there is TMP_DIR<a name="line.456"></a>
<span class="sourceLineNo">457</span> assertTrue(fs.exists(tmpPath));<a name="line.457"></a>
<span class="sourceLineNo">458</span> // TMP_DIR should have been cleaned-up<a name="line.458"></a>
<span class="sourceLineNo">459</span> assertNull(BulkLoadHFilesTool.TMP_DIR + " should be empty.",<a name="line.459"></a>
<span class="sourceLineNo">460</span> CommonFSUtils.listStatus(fs, tmpPath));<a name="line.460"></a>
<span class="sourceLineNo">461</span> assertExpectedTable(util.getConnection(), table, ROWCOUNT, 2);<a name="line.461"></a>
<span class="sourceLineNo">462</span> }<a name="line.462"></a>
<span class="sourceLineNo">463</span><a name="line.463"></a>
<span class="sourceLineNo">464</span> /**<a name="line.464"></a>
<span class="sourceLineNo">465</span> * This simulates an remote exception which should cause LIHF to exit with an exception.<a name="line.465"></a>
<span class="sourceLineNo">466</span> */<a name="line.466"></a>
<span class="sourceLineNo">467</span> @Test(expected = IOException.class)<a name="line.467"></a>
<span class="sourceLineNo">468</span> public void testGroupOrSplitFailure() throws Exception {<a name="line.468"></a>
<span class="sourceLineNo">469</span> final TableName tableName = TableName.valueOf(name.getMethodName());<a name="line.469"></a>
<span class="sourceLineNo">470</span> setupTable(util.getConnection(), tableName, 10);<a name="line.470"></a>
<span class="sourceLineNo">471</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(util.getConfiguration()) {<a name="line.471"></a>
<span class="sourceLineNo">472</span><a name="line.472"></a>
<span class="sourceLineNo">473</span> private int i = 0;<a name="line.473"></a>
<span class="sourceLineNo">474</span><a name="line.474"></a>
<span class="sourceLineNo">475</span> @Override<a name="line.475"></a>
<span class="sourceLineNo">476</span> protected Pair&lt;List&lt;LoadQueueItem&gt;, String&gt; groupOrSplit(AsyncClusterConnection conn,<a name="line.476"></a>
<span class="sourceLineNo">477</span> TableName tableName, Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups, LoadQueueItem item,<a name="line.477"></a>
<span class="sourceLineNo">478</span> List&lt;Pair&lt;byte[], byte[]&gt;&gt; startEndKeys) throws IOException {<a name="line.478"></a>
<span class="sourceLineNo">479</span> i++;<a name="line.479"></a>
<span class="sourceLineNo">480</span><a name="line.480"></a>
<span class="sourceLineNo">481</span> if (i == 5) {<a name="line.481"></a>
<span class="sourceLineNo">482</span> throw new IOException("failure");<a name="line.482"></a>
<span class="sourceLineNo">483</span> }<a name="line.483"></a>
<span class="sourceLineNo">484</span> return super.groupOrSplit(conn, tableName, regionGroups, item, startEndKeys);<a name="line.484"></a>
<span class="sourceLineNo">485</span> }<a name="line.485"></a>
<span class="sourceLineNo">486</span> };<a name="line.486"></a>
<span class="sourceLineNo">487</span><a name="line.487"></a>
<span class="sourceLineNo">488</span> // create HFiles for different column families<a name="line.488"></a>
<span class="sourceLineNo">489</span> Path dir = buildBulkFiles(tableName, 1);<a name="line.489"></a>
<span class="sourceLineNo">490</span> loader.bulkLoad(tableName, dir);<a name="line.490"></a>
<span class="sourceLineNo">491</span> }<a name="line.491"></a>
<span class="sourceLineNo">492</span><a name="line.492"></a>
<span class="sourceLineNo">493</span> /**<a name="line.493"></a>
<span class="sourceLineNo">494</span> * We are testing a split after initial validation but before the atomic bulk load call. We cannot<a name="line.494"></a>
<span class="sourceLineNo">495</span> * use presplitting to test this path, so we actually inject a split just before the atomic region<a name="line.495"></a>
<span class="sourceLineNo">496</span> * load. However, we will pass null item2RegionMap and that should not affect the bulk load<a name="line.496"></a>
<span class="sourceLineNo">497</span> * behavior.<a name="line.497"></a>
<span class="sourceLineNo">498</span> */<a name="line.498"></a>
<span class="sourceLineNo">499</span> @Test<a name="line.499"></a>
<span class="sourceLineNo">500</span> public void testSplitWhileBulkLoadPhaseWithoutItemMap() throws Exception {<a name="line.500"></a>
<span class="sourceLineNo">501</span> final TableName table = TableName.valueOf(name.getMethodName());<a name="line.501"></a>
<span class="sourceLineNo">502</span> setupTable(util.getConnection(), table, 10);<a name="line.502"></a>
<span class="sourceLineNo">503</span> populateTable(util.getConnection(), table, 1);<a name="line.503"></a>
<span class="sourceLineNo">504</span> assertExpectedTable(table, ROWCOUNT, 1);<a name="line.504"></a>
<span class="sourceLineNo">505</span><a name="line.505"></a>
<span class="sourceLineNo">506</span> // Now let's cause trouble. This will occur after checks and cause bulk<a name="line.506"></a>
<span class="sourceLineNo">507</span> // files to fail when attempt to atomically import. This is recoverable.<a name="line.507"></a>
<span class="sourceLineNo">508</span> final AtomicInteger attemptedCalls = new AtomicInteger();<a name="line.508"></a>
<span class="sourceLineNo">509</span> BulkLoadHFilesTool loader = new BulkLoadHFilesTool(util.getConfiguration()) {<a name="line.509"></a>
<span class="sourceLineNo">510</span><a name="line.510"></a>
<span class="sourceLineNo">511</span> @Override<a name="line.511"></a>
<span class="sourceLineNo">512</span> protected void bulkLoadPhase(final AsyncClusterConnection conn, final TableName tableName,<a name="line.512"></a>
<span class="sourceLineNo">513</span> final Deque&lt;LoadQueueItem&gt; queue, final Multimap&lt;ByteBuffer, LoadQueueItem&gt; regionGroups,<a name="line.513"></a>
<span class="sourceLineNo">514</span> final boolean copyFiles, final Map&lt;LoadQueueItem, ByteBuffer&gt; item2RegionMap)<a name="line.514"></a>
<span class="sourceLineNo">515</span> throws IOException {<a name="line.515"></a>
<span class="sourceLineNo">516</span><a name="line.516"></a>
<span class="sourceLineNo">517</span> int i = attemptedCalls.incrementAndGet();<a name="line.517"></a>
<span class="sourceLineNo">518</span> if (i == 1) {<a name="line.518"></a>
<span class="sourceLineNo">519</span> // On first attempt force a split.<a name="line.519"></a>
<span class="sourceLineNo">520</span> forceSplit(table);<a name="line.520"></a>
<span class="sourceLineNo">521</span> }<a name="line.521"></a>
<span class="sourceLineNo">522</span><a name="line.522"></a>
<span class="sourceLineNo">523</span> // Passing item2RegionMap null<a name="line.523"></a>
<span class="sourceLineNo">524</span> // In the absence of LoadQueueItem, bulk load should work as expected<a name="line.524"></a>
<span class="sourceLineNo">525</span> super.bulkLoadPhase(conn, tableName, queue, regionGroups, copyFiles, null);<a name="line.525"></a>
<span class="sourceLineNo">526</span> }<a name="line.526"></a>
<span class="sourceLineNo">527</span><a name="line.527"></a>
<span class="sourceLineNo">528</span> };<a name="line.528"></a>
<span class="sourceLineNo">529</span><a name="line.529"></a>
<span class="sourceLineNo">530</span> // create HFiles for different column families<a name="line.530"></a>
<span class="sourceLineNo">531</span> Path dir = buildBulkFiles(table, 2);<a name="line.531"></a>
<span class="sourceLineNo">532</span> loader.bulkLoad(table, dir);<a name="line.532"></a>
<span class="sourceLineNo">533</span><a name="line.533"></a>
<span class="sourceLineNo">534</span> // check that data was loaded<a name="line.534"></a>
<span class="sourceLineNo">535</span> // The three expected attempts are 1) failure because need to split, 2)<a name="line.535"></a>
<span class="sourceLineNo">536</span> // load of split top 3) load of split bottom<a name="line.536"></a>
<span class="sourceLineNo">537</span> assertEquals(3, attemptedCalls.get());<a name="line.537"></a>
<span class="sourceLineNo">538</span> assertExpectedTable(table, ROWCOUNT, 2);<a name="line.538"></a>
<span class="sourceLineNo">539</span> }<a name="line.539"></a>
<span class="sourceLineNo">540</span><a name="line.540"></a>
<span class="sourceLineNo">541</span> /**<a name="line.541"></a>
<span class="sourceLineNo">542</span> * Checks that all columns have the expected value and that there is the expected number of rows.<a name="line.542"></a>
<span class="sourceLineNo">543</span> */<a name="line.543"></a>
<span class="sourceLineNo">544</span> void assertExpectedTable(final Connection connection, TableName table, int count, int value)<a name="line.544"></a>
<span class="sourceLineNo">545</span> throws IOException {<a name="line.545"></a>
<span class="sourceLineNo">546</span> TableDescriptor htd = util.getAdmin().getDescriptor(table);<a name="line.546"></a>
<span class="sourceLineNo">547</span> assertNotNull(htd);<a name="line.547"></a>
<span class="sourceLineNo">548</span> try (Table t = connection.getTable(table); ResultScanner sr = t.getScanner(new Scan())) {<a name="line.548"></a>
<span class="sourceLineNo">549</span> int i = 0;<a name="line.549"></a>
<span class="sourceLineNo">550</span> for (Result r; (r = sr.next()) != null;) {<a name="line.550"></a>
<span class="sourceLineNo">551</span> r.getNoVersionMap().values().stream().flatMap(m -&gt; m.values().stream())<a name="line.551"></a>
<span class="sourceLineNo">552</span> .forEach(v -&gt; assertArrayEquals(value(value), v));<a name="line.552"></a>
<span class="sourceLineNo">553</span> i++;<a name="line.553"></a>
<span class="sourceLineNo">554</span> }<a name="line.554"></a>
<span class="sourceLineNo">555</span> assertEquals(count, i);<a name="line.555"></a>
<span class="sourceLineNo">556</span> } catch (IOException e) {<a name="line.556"></a>
<span class="sourceLineNo">557</span> fail("Failed due to exception");<a name="line.557"></a>
<span class="sourceLineNo">558</span> }<a name="line.558"></a>
<span class="sourceLineNo">559</span> }<a name="line.559"></a>
<span class="sourceLineNo">560</span>}<a name="line.560"></a>
</pre>
</div>
</body>
</html>