blob: 121ce9224b63cc52fb93664c81d58a39a565f782 [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.wal;<a name="line.18"></a>
<span class="sourceLineNo">019</span><a name="line.19"></a>
<span class="sourceLineNo">020</span>import java.io.EOFException;<a name="line.20"></a>
<span class="sourceLineNo">021</span>import java.io.FileNotFoundException;<a name="line.21"></a>
<span class="sourceLineNo">022</span>import java.io.IOException;<a name="line.22"></a>
<span class="sourceLineNo">023</span>import java.io.InterruptedIOException;<a name="line.23"></a>
<span class="sourceLineNo">024</span>import java.text.ParseException;<a name="line.24"></a>
<span class="sourceLineNo">025</span>import java.util.ArrayList;<a name="line.25"></a>
<span class="sourceLineNo">026</span>import java.util.Collections;<a name="line.26"></a>
<span class="sourceLineNo">027</span>import java.util.List;<a name="line.27"></a>
<span class="sourceLineNo">028</span>import java.util.Map;<a name="line.28"></a>
<span class="sourceLineNo">029</span>import java.util.TreeMap;<a name="line.29"></a>
<span class="sourceLineNo">030</span>import java.util.concurrent.ConcurrentHashMap;<a name="line.30"></a>
<span class="sourceLineNo">031</span>import java.util.concurrent.atomic.AtomicReference;<a name="line.31"></a>
<span class="sourceLineNo">032</span>import org.apache.hadoop.conf.Configuration;<a name="line.32"></a>
<span class="sourceLineNo">033</span>import org.apache.hadoop.fs.FileStatus;<a name="line.33"></a>
<span class="sourceLineNo">034</span>import org.apache.hadoop.fs.FileSystem;<a name="line.34"></a>
<span class="sourceLineNo">035</span>import org.apache.hadoop.fs.Path;<a name="line.35"></a>
<span class="sourceLineNo">036</span>import org.apache.hadoop.hbase.HBaseConfiguration;<a name="line.36"></a>
<span class="sourceLineNo">037</span>import org.apache.hadoop.hbase.HConstants;<a name="line.37"></a>
<span class="sourceLineNo">038</span>import org.apache.hadoop.hbase.TableName;<a name="line.38"></a>
<span class="sourceLineNo">039</span>import org.apache.hadoop.hbase.coordination.SplitLogWorkerCoordination;<a name="line.39"></a>
<span class="sourceLineNo">040</span>import org.apache.hadoop.hbase.master.SplitLogManager;<a name="line.40"></a>
<span class="sourceLineNo">041</span>import org.apache.hadoop.hbase.monitoring.MonitoredTask;<a name="line.41"></a>
<span class="sourceLineNo">042</span>import org.apache.hadoop.hbase.monitoring.TaskMonitor;<a name="line.42"></a>
<span class="sourceLineNo">043</span>import org.apache.hadoop.hbase.procedure2.util.StringUtils;<a name="line.43"></a>
<span class="sourceLineNo">044</span>import org.apache.hadoop.hbase.regionserver.LastSequenceId;<a name="line.44"></a>
<span class="sourceLineNo">045</span>import org.apache.hadoop.hbase.regionserver.RegionServerServices;<a name="line.45"></a>
<span class="sourceLineNo">046</span>import org.apache.hadoop.hbase.regionserver.wal.WALCellCodec;<a name="line.46"></a>
<span class="sourceLineNo">047</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.47"></a>
<span class="sourceLineNo">048</span>import org.apache.hadoop.hbase.util.CancelableProgressable;<a name="line.48"></a>
<span class="sourceLineNo">049</span>import org.apache.hadoop.hbase.util.CommonFSUtils;<a name="line.49"></a>
<span class="sourceLineNo">050</span>import org.apache.hadoop.hbase.util.EnvironmentEdgeManager;<a name="line.50"></a>
<span class="sourceLineNo">051</span>import org.apache.hadoop.hbase.util.RecoverLeaseFSUtils;<a name="line.51"></a>
<span class="sourceLineNo">052</span>import org.apache.hadoop.hbase.wal.WAL.Entry;<a name="line.52"></a>
<span class="sourceLineNo">053</span>import org.apache.hadoop.hbase.wal.WAL.Reader;<a name="line.53"></a>
<span class="sourceLineNo">054</span>import org.apache.hadoop.ipc.RemoteException;<a name="line.54"></a>
<span class="sourceLineNo">055</span>import org.apache.yetus.audience.InterfaceAudience;<a name="line.55"></a>
<span class="sourceLineNo">056</span>import org.slf4j.Logger;<a name="line.56"></a>
<span class="sourceLineNo">057</span>import org.slf4j.LoggerFactory;<a name="line.57"></a>
<span class="sourceLineNo">058</span><a name="line.58"></a>
<span class="sourceLineNo">059</span>import org.apache.hbase.thirdparty.com.google.common.base.Preconditions;<a name="line.59"></a>
<span class="sourceLineNo">060</span>import org.apache.hbase.thirdparty.com.google.protobuf.TextFormat;<a name="line.60"></a>
<span class="sourceLineNo">061</span><a name="line.61"></a>
<span class="sourceLineNo">062</span>import org.apache.hadoop.hbase.shaded.protobuf.generated.ClusterStatusProtos.RegionStoreSequenceIds;<a name="line.62"></a>
<span class="sourceLineNo">063</span>import org.apache.hadoop.hbase.shaded.protobuf.generated.ClusterStatusProtos.StoreSequenceId;<a name="line.63"></a>
<span class="sourceLineNo">064</span><a name="line.64"></a>
<span class="sourceLineNo">065</span>/**<a name="line.65"></a>
<span class="sourceLineNo">066</span> * Split RegionServer WAL files. Splits the WAL into new files, one per region, to be picked up on<a name="line.66"></a>
<span class="sourceLineNo">067</span> * Region reopen. Deletes the split WAL when finished. Create an instance and call<a name="line.67"></a>
<span class="sourceLineNo">068</span> * {@link #splitWAL(FileStatus, CancelableProgressable)} per file or use static helper methods.<a name="line.68"></a>
<span class="sourceLineNo">069</span> */<a name="line.69"></a>
<span class="sourceLineNo">070</span>@InterfaceAudience.Private<a name="line.70"></a>
<span class="sourceLineNo">071</span>public class WALSplitter {<a name="line.71"></a>
<span class="sourceLineNo">072</span> private static final Logger LOG = LoggerFactory.getLogger(WALSplitter.class);<a name="line.72"></a>
<span class="sourceLineNo">073</span> public static final String SPLIT_SKIP_ERRORS_KEY = "hbase.hlog.split.skip.errors";<a name="line.73"></a>
<span class="sourceLineNo">074</span><a name="line.74"></a>
<span class="sourceLineNo">075</span> /**<a name="line.75"></a>
<span class="sourceLineNo">076</span> * By default we retry errors in splitting, rather than skipping.<a name="line.76"></a>
<span class="sourceLineNo">077</span> */<a name="line.77"></a>
<span class="sourceLineNo">078</span> public static final boolean SPLIT_SKIP_ERRORS_DEFAULT = false;<a name="line.78"></a>
<span class="sourceLineNo">079</span><a name="line.79"></a>
<span class="sourceLineNo">080</span> // Parameters for split process<a name="line.80"></a>
<span class="sourceLineNo">081</span> protected final Path walRootDir;<a name="line.81"></a>
<span class="sourceLineNo">082</span> protected final FileSystem walFS;<a name="line.82"></a>
<span class="sourceLineNo">083</span> protected final Configuration conf;<a name="line.83"></a>
<span class="sourceLineNo">084</span> final Path rootDir;<a name="line.84"></a>
<span class="sourceLineNo">085</span> final FileSystem rootFS;<a name="line.85"></a>
<span class="sourceLineNo">086</span> final RegionServerServices rsServices;<a name="line.86"></a>
<span class="sourceLineNo">087</span><a name="line.87"></a>
<span class="sourceLineNo">088</span> // Major subcomponents of the split process.<a name="line.88"></a>
<span class="sourceLineNo">089</span> // These are separated into inner classes to make testing easier.<a name="line.89"></a>
<span class="sourceLineNo">090</span> OutputSink outputSink;<a name="line.90"></a>
<span class="sourceLineNo">091</span> private EntryBuffers entryBuffers;<a name="line.91"></a>
<span class="sourceLineNo">092</span><a name="line.92"></a>
<span class="sourceLineNo">093</span> /**<a name="line.93"></a>
<span class="sourceLineNo">094</span> * Coordinator for split log. Used by the zk-based log splitter. Not used by the procedure<a name="line.94"></a>
<span class="sourceLineNo">095</span> * v2-based log splitter.<a name="line.95"></a>
<span class="sourceLineNo">096</span> */<a name="line.96"></a>
<span class="sourceLineNo">097</span> private SplitLogWorkerCoordination splitLogWorkerCoordination;<a name="line.97"></a>
<span class="sourceLineNo">098</span><a name="line.98"></a>
<span class="sourceLineNo">099</span> private final WALFactory walFactory;<a name="line.99"></a>
<span class="sourceLineNo">100</span><a name="line.100"></a>
<span class="sourceLineNo">101</span> // For checking the latest flushed sequence id<a name="line.101"></a>
<span class="sourceLineNo">102</span> protected final LastSequenceId sequenceIdChecker;<a name="line.102"></a>
<span class="sourceLineNo">103</span><a name="line.103"></a>
<span class="sourceLineNo">104</span> // Map encodedRegionName -&gt; lastFlushedSequenceId<a name="line.104"></a>
<span class="sourceLineNo">105</span> protected Map&lt;String, Long&gt; lastFlushedSequenceIds = new ConcurrentHashMap&lt;&gt;();<a name="line.105"></a>
<span class="sourceLineNo">106</span><a name="line.106"></a>
<span class="sourceLineNo">107</span> // Map encodedRegionName -&gt; maxSeqIdInStores<a name="line.107"></a>
<span class="sourceLineNo">108</span> protected Map&lt;String, Map&lt;byte[], Long&gt;&gt; regionMaxSeqIdInStores = new ConcurrentHashMap&lt;&gt;();<a name="line.108"></a>
<span class="sourceLineNo">109</span><a name="line.109"></a>
<span class="sourceLineNo">110</span> // the file being split currently<a name="line.110"></a>
<span class="sourceLineNo">111</span> private FileStatus fileBeingSplit;<a name="line.111"></a>
<span class="sourceLineNo">112</span><a name="line.112"></a>
<span class="sourceLineNo">113</span> private final String tmpDirName;<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> * Split WAL directly to hfiles instead of into intermediary 'recovered.edits' files.<a name="line.116"></a>
<span class="sourceLineNo">117</span> */<a name="line.117"></a>
<span class="sourceLineNo">118</span> public static final String WAL_SPLIT_TO_HFILE = "hbase.wal.split.to.hfile";<a name="line.118"></a>
<span class="sourceLineNo">119</span> public static final boolean DEFAULT_WAL_SPLIT_TO_HFILE = false;<a name="line.119"></a>
<span class="sourceLineNo">120</span><a name="line.120"></a>
<span class="sourceLineNo">121</span> /**<a name="line.121"></a>
<span class="sourceLineNo">122</span> * True if we are to run with bounded amount of writers rather than let the count blossom. Default<a name="line.122"></a>
<span class="sourceLineNo">123</span> * is 'false'. Does not apply if you have set 'hbase.wal.split.to.hfile' as that is always<a name="line.123"></a>
<span class="sourceLineNo">124</span> * bounded. Only applies when you are doing recovery to 'recovered.edits' files (the old default).<a name="line.124"></a>
<span class="sourceLineNo">125</span> * Bounded writing tends to have higher throughput.<a name="line.125"></a>
<span class="sourceLineNo">126</span> */<a name="line.126"></a>
<span class="sourceLineNo">127</span> public final static String SPLIT_WRITER_CREATION_BOUNDED = "hbase.split.writer.creation.bounded";<a name="line.127"></a>
<span class="sourceLineNo">128</span><a name="line.128"></a>
<span class="sourceLineNo">129</span> public final static String SPLIT_WAL_BUFFER_SIZE = "hbase.regionserver.hlog.splitlog.buffersize";<a name="line.129"></a>
<span class="sourceLineNo">130</span> public final static String SPLIT_WAL_WRITER_THREADS =<a name="line.130"></a>
<span class="sourceLineNo">131</span> "hbase.regionserver.hlog.splitlog.writer.threads";<a name="line.131"></a>
<span class="sourceLineNo">132</span><a name="line.132"></a>
<span class="sourceLineNo">133</span> private final int numWriterThreads;<a name="line.133"></a>
<span class="sourceLineNo">134</span> private final long bufferSize;<a name="line.134"></a>
<span class="sourceLineNo">135</span> private final boolean splitWriterCreationBounded;<a name="line.135"></a>
<span class="sourceLineNo">136</span> private final boolean hfile;<a name="line.136"></a>
<span class="sourceLineNo">137</span> private final boolean skipErrors;<a name="line.137"></a>
<span class="sourceLineNo">138</span><a name="line.138"></a>
<span class="sourceLineNo">139</span> WALSplitter(final WALFactory factory, Configuration conf, Path walRootDir, FileSystem walFS,<a name="line.139"></a>
<span class="sourceLineNo">140</span> Path rootDir, FileSystem rootFS) {<a name="line.140"></a>
<span class="sourceLineNo">141</span> this(factory, conf, walRootDir, walFS, rootDir, rootFS, null, null, null);<a name="line.141"></a>
<span class="sourceLineNo">142</span> }<a name="line.142"></a>
<span class="sourceLineNo">143</span><a name="line.143"></a>
<span class="sourceLineNo">144</span> WALSplitter(final WALFactory factory, Configuration conf, Path walRootDir, FileSystem walFS,<a name="line.144"></a>
<span class="sourceLineNo">145</span> Path rootDir, FileSystem rootFS, LastSequenceId idChecker,<a name="line.145"></a>
<span class="sourceLineNo">146</span> SplitLogWorkerCoordination splitLogWorkerCoordination, RegionServerServices rsServices) {<a name="line.146"></a>
<span class="sourceLineNo">147</span> this.conf = HBaseConfiguration.create(conf);<a name="line.147"></a>
<span class="sourceLineNo">148</span> String codecClassName =<a name="line.148"></a>
<span class="sourceLineNo">149</span> conf.get(WALCellCodec.WAL_CELL_CODEC_CLASS_KEY, WALCellCodec.class.getName());<a name="line.149"></a>
<span class="sourceLineNo">150</span> this.conf.set(HConstants.RPC_CODEC_CONF_KEY, codecClassName);<a name="line.150"></a>
<span class="sourceLineNo">151</span> this.walRootDir = walRootDir;<a name="line.151"></a>
<span class="sourceLineNo">152</span> this.walFS = walFS;<a name="line.152"></a>
<span class="sourceLineNo">153</span> this.rootDir = rootDir;<a name="line.153"></a>
<span class="sourceLineNo">154</span> this.rootFS = rootFS;<a name="line.154"></a>
<span class="sourceLineNo">155</span> this.sequenceIdChecker = idChecker;<a name="line.155"></a>
<span class="sourceLineNo">156</span> this.splitLogWorkerCoordination = splitLogWorkerCoordination;<a name="line.156"></a>
<span class="sourceLineNo">157</span> this.rsServices = rsServices;<a name="line.157"></a>
<span class="sourceLineNo">158</span> this.walFactory = factory;<a name="line.158"></a>
<span class="sourceLineNo">159</span> this.tmpDirName =<a name="line.159"></a>
<span class="sourceLineNo">160</span> conf.get(HConstants.TEMPORARY_FS_DIRECTORY_KEY, HConstants.DEFAULT_TEMPORARY_HDFS_DIRECTORY);<a name="line.160"></a>
<span class="sourceLineNo">161</span> // if we limit the number of writers opened for sinking recovered edits<a name="line.161"></a>
<span class="sourceLineNo">162</span> this.splitWriterCreationBounded = conf.getBoolean(SPLIT_WRITER_CREATION_BOUNDED, false);<a name="line.162"></a>
<span class="sourceLineNo">163</span> this.bufferSize = this.conf.getLong(SPLIT_WAL_BUFFER_SIZE, 128 * 1024 * 1024);<a name="line.163"></a>
<span class="sourceLineNo">164</span> this.numWriterThreads = this.conf.getInt(SPLIT_WAL_WRITER_THREADS, 3);<a name="line.164"></a>
<span class="sourceLineNo">165</span> this.hfile = conf.getBoolean(WAL_SPLIT_TO_HFILE, DEFAULT_WAL_SPLIT_TO_HFILE);<a name="line.165"></a>
<span class="sourceLineNo">166</span> this.skipErrors = conf.getBoolean(SPLIT_SKIP_ERRORS_KEY, SPLIT_SKIP_ERRORS_DEFAULT);<a name="line.166"></a>
<span class="sourceLineNo">167</span> }<a name="line.167"></a>
<span class="sourceLineNo">168</span><a name="line.168"></a>
<span class="sourceLineNo">169</span> WALFactory getWalFactory() {<a name="line.169"></a>
<span class="sourceLineNo">170</span> return this.walFactory;<a name="line.170"></a>
<span class="sourceLineNo">171</span> }<a name="line.171"></a>
<span class="sourceLineNo">172</span><a name="line.172"></a>
<span class="sourceLineNo">173</span> FileStatus getFileBeingSplit() {<a name="line.173"></a>
<span class="sourceLineNo">174</span> return fileBeingSplit;<a name="line.174"></a>
<span class="sourceLineNo">175</span> }<a name="line.175"></a>
<span class="sourceLineNo">176</span><a name="line.176"></a>
<span class="sourceLineNo">177</span> String getTmpDirName() {<a name="line.177"></a>
<span class="sourceLineNo">178</span> return this.tmpDirName;<a name="line.178"></a>
<span class="sourceLineNo">179</span> }<a name="line.179"></a>
<span class="sourceLineNo">180</span><a name="line.180"></a>
<span class="sourceLineNo">181</span> Map&lt;String, Map&lt;byte[], Long&gt;&gt; getRegionMaxSeqIdInStores() {<a name="line.181"></a>
<span class="sourceLineNo">182</span> return regionMaxSeqIdInStores;<a name="line.182"></a>
<span class="sourceLineNo">183</span> }<a name="line.183"></a>
<span class="sourceLineNo">184</span><a name="line.184"></a>
<span class="sourceLineNo">185</span> /**<a name="line.185"></a>
<span class="sourceLineNo">186</span> * Splits a WAL file. Used by old {@link org.apache.hadoop.hbase.regionserver.SplitLogWorker} and<a name="line.186"></a>
<span class="sourceLineNo">187</span> * tests. Not used by new procedure-based WAL splitter.<a name="line.187"></a>
<span class="sourceLineNo">188</span> * @return false if it is interrupted by the progress-able.<a name="line.188"></a>
<span class="sourceLineNo">189</span> */<a name="line.189"></a>
<span class="sourceLineNo">190</span> public static boolean splitLogFile(Path walDir, FileStatus logfile, FileSystem walFS,<a name="line.190"></a>
<span class="sourceLineNo">191</span> Configuration conf, CancelableProgressable reporter, LastSequenceId idChecker,<a name="line.191"></a>
<span class="sourceLineNo">192</span> SplitLogWorkerCoordination splitLogWorkerCoordination, WALFactory factory,<a name="line.192"></a>
<span class="sourceLineNo">193</span> RegionServerServices rsServices) throws IOException {<a name="line.193"></a>
<span class="sourceLineNo">194</span> Path rootDir = CommonFSUtils.getRootDir(conf);<a name="line.194"></a>
<span class="sourceLineNo">195</span> FileSystem rootFS = rootDir.getFileSystem(conf);<a name="line.195"></a>
<span class="sourceLineNo">196</span> WALSplitter splitter = new WALSplitter(factory, conf, walDir, walFS, rootDir, rootFS, idChecker,<a name="line.196"></a>
<span class="sourceLineNo">197</span> splitLogWorkerCoordination, rsServices);<a name="line.197"></a>
<span class="sourceLineNo">198</span> // splitWAL returns a data structure with whether split is finished and if the file is corrupt.<a name="line.198"></a>
<span class="sourceLineNo">199</span> // We don't need to propagate corruption flag here because it is propagated by the<a name="line.199"></a>
<span class="sourceLineNo">200</span> // SplitLogWorkerCoordination.<a name="line.200"></a>
<span class="sourceLineNo">201</span> return splitter.splitWAL(logfile, reporter).isFinished();<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> /**<a name="line.204"></a>
<span class="sourceLineNo">205</span> * Split a folder of WAL files. Delete the directory when done. Used by tools and unit tests. It<a name="line.205"></a>
<span class="sourceLineNo">206</span> * should be package private. It is public only because TestWALObserver is in a different package,<a name="line.206"></a>
<span class="sourceLineNo">207</span> * which uses this method to do log splitting.<a name="line.207"></a>
<span class="sourceLineNo">208</span> * @return List of output files created by the split.<a name="line.208"></a>
<span class="sourceLineNo">209</span> */<a name="line.209"></a>
<span class="sourceLineNo">210</span> public static List&lt;Path&gt; split(Path walRootDir, Path walsDir, Path archiveDir, FileSystem walFS,<a name="line.210"></a>
<span class="sourceLineNo">211</span> Configuration conf, final WALFactory factory) throws IOException {<a name="line.211"></a>
<span class="sourceLineNo">212</span> Path rootDir = CommonFSUtils.getRootDir(conf);<a name="line.212"></a>
<span class="sourceLineNo">213</span> FileSystem rootFS = rootDir.getFileSystem(conf);<a name="line.213"></a>
<span class="sourceLineNo">214</span> WALSplitter splitter = new WALSplitter(factory, conf, walRootDir, walFS, rootDir, rootFS);<a name="line.214"></a>
<span class="sourceLineNo">215</span> final List&lt;FileStatus&gt; wals =<a name="line.215"></a>
<span class="sourceLineNo">216</span> SplitLogManager.getFileList(conf, Collections.singletonList(walsDir), null);<a name="line.216"></a>
<span class="sourceLineNo">217</span> List&lt;Path&gt; splits = new ArrayList&lt;&gt;();<a name="line.217"></a>
<span class="sourceLineNo">218</span> if (!wals.isEmpty()) {<a name="line.218"></a>
<span class="sourceLineNo">219</span> for (FileStatus wal : wals) {<a name="line.219"></a>
<span class="sourceLineNo">220</span> SplitWALResult splitWALResult = splitter.splitWAL(wal, null);<a name="line.220"></a>
<span class="sourceLineNo">221</span> if (splitWALResult.isFinished()) {<a name="line.221"></a>
<span class="sourceLineNo">222</span> WALSplitUtil.archive(wal.getPath(), splitWALResult.isCorrupt(), archiveDir, walFS, conf);<a name="line.222"></a>
<span class="sourceLineNo">223</span> // splitter.outputSink.splits is mark as final, do not need null check<a name="line.223"></a>
<span class="sourceLineNo">224</span> splits.addAll(splitter.outputSink.splits);<a name="line.224"></a>
<span class="sourceLineNo">225</span> }<a name="line.225"></a>
<span class="sourceLineNo">226</span> }<a name="line.226"></a>
<span class="sourceLineNo">227</span> }<a name="line.227"></a>
<span class="sourceLineNo">228</span> if (!walFS.delete(walsDir, true)) {<a name="line.228"></a>
<span class="sourceLineNo">229</span> throw new IOException("Unable to delete src dir " + walsDir);<a name="line.229"></a>
<span class="sourceLineNo">230</span> }<a name="line.230"></a>
<span class="sourceLineNo">231</span> return splits;<a name="line.231"></a>
<span class="sourceLineNo">232</span> }<a name="line.232"></a>
<span class="sourceLineNo">233</span><a name="line.233"></a>
<span class="sourceLineNo">234</span> /**<a name="line.234"></a>
<span class="sourceLineNo">235</span> * Data structure returned as result by #splitWAL(FileStatus, CancelableProgressable). Test<a name="line.235"></a>
<span class="sourceLineNo">236</span> * {@link #isFinished()} to see if we are done with the WAL and {@link #isCorrupt()} for if the<a name="line.236"></a>
<span class="sourceLineNo">237</span> * WAL is corrupt.<a name="line.237"></a>
<span class="sourceLineNo">238</span> */<a name="line.238"></a>
<span class="sourceLineNo">239</span> static final class SplitWALResult {<a name="line.239"></a>
<span class="sourceLineNo">240</span> private final boolean finished;<a name="line.240"></a>
<span class="sourceLineNo">241</span> private final boolean corrupt;<a name="line.241"></a>
<span class="sourceLineNo">242</span><a name="line.242"></a>
<span class="sourceLineNo">243</span> private SplitWALResult(boolean finished, boolean corrupt) {<a name="line.243"></a>
<span class="sourceLineNo">244</span> this.finished = finished;<a name="line.244"></a>
<span class="sourceLineNo">245</span> this.corrupt = corrupt;<a name="line.245"></a>
<span class="sourceLineNo">246</span> }<a name="line.246"></a>
<span class="sourceLineNo">247</span><a name="line.247"></a>
<span class="sourceLineNo">248</span> public boolean isFinished() {<a name="line.248"></a>
<span class="sourceLineNo">249</span> return finished;<a name="line.249"></a>
<span class="sourceLineNo">250</span> }<a name="line.250"></a>
<span class="sourceLineNo">251</span><a name="line.251"></a>
<span class="sourceLineNo">252</span> public boolean isCorrupt() {<a name="line.252"></a>
<span class="sourceLineNo">253</span> return corrupt;<a name="line.253"></a>
<span class="sourceLineNo">254</span> }<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> * Setup the output sinks and entry buffers ahead of splitting WAL.<a name="line.258"></a>
<span class="sourceLineNo">259</span> */<a name="line.259"></a>
<span class="sourceLineNo">260</span> private void createOutputSinkAndEntryBuffers() {<a name="line.260"></a>
<span class="sourceLineNo">261</span> PipelineController controller = new PipelineController();<a name="line.261"></a>
<span class="sourceLineNo">262</span> if (this.hfile) {<a name="line.262"></a>
<span class="sourceLineNo">263</span> this.entryBuffers = new BoundedEntryBuffers(controller, this.bufferSize);<a name="line.263"></a>
<span class="sourceLineNo">264</span> this.outputSink = new BoundedRecoveredHFilesOutputSink(this, controller, this.entryBuffers,<a name="line.264"></a>
<span class="sourceLineNo">265</span> this.numWriterThreads);<a name="line.265"></a>
<span class="sourceLineNo">266</span> } else if (this.splitWriterCreationBounded) {<a name="line.266"></a>
<span class="sourceLineNo">267</span> this.entryBuffers = new BoundedEntryBuffers(controller, this.bufferSize);<a name="line.267"></a>
<span class="sourceLineNo">268</span> this.outputSink = new BoundedRecoveredEditsOutputSink(this, controller, this.entryBuffers,<a name="line.268"></a>
<span class="sourceLineNo">269</span> this.numWriterThreads);<a name="line.269"></a>
<span class="sourceLineNo">270</span> } else {<a name="line.270"></a>
<span class="sourceLineNo">271</span> this.entryBuffers = new EntryBuffers(controller, this.bufferSize);<a name="line.271"></a>
<span class="sourceLineNo">272</span> this.outputSink =<a name="line.272"></a>
<span class="sourceLineNo">273</span> new RecoveredEditsOutputSink(this, controller, this.entryBuffers, this.numWriterThreads);<a name="line.273"></a>
<span class="sourceLineNo">274</span> }<a name="line.274"></a>
<span class="sourceLineNo">275</span> }<a name="line.275"></a>
<span class="sourceLineNo">276</span><a name="line.276"></a>
<span class="sourceLineNo">277</span> /**<a name="line.277"></a>
<span class="sourceLineNo">278</span> * WAL splitting implementation, splits one WAL file.<a name="line.278"></a>
<span class="sourceLineNo">279</span> * @param walStatus should be for an actual WAL file.<a name="line.279"></a>
<span class="sourceLineNo">280</span> */<a name="line.280"></a>
<span class="sourceLineNo">281</span> SplitWALResult splitWAL(FileStatus walStatus, CancelableProgressable cancel) throws IOException {<a name="line.281"></a>
<span class="sourceLineNo">282</span> Path wal = walStatus.getPath();<a name="line.282"></a>
<span class="sourceLineNo">283</span> Preconditions.checkArgument(walStatus.isFile(), "Not a regular file " + wal.toString());<a name="line.283"></a>
<span class="sourceLineNo">284</span> boolean corrupt = false;<a name="line.284"></a>
<span class="sourceLineNo">285</span> int interval = conf.getInt("hbase.splitlog.report.interval.loglines", 1024);<a name="line.285"></a>
<span class="sourceLineNo">286</span> boolean outputSinkStarted = false;<a name="line.286"></a>
<span class="sourceLineNo">287</span> boolean cancelled = false;<a name="line.287"></a>
<span class="sourceLineNo">288</span> int editsCount = 0;<a name="line.288"></a>
<span class="sourceLineNo">289</span> int editsSkipped = 0;<a name="line.289"></a>
<span class="sourceLineNo">290</span> MonitoredTask status = TaskMonitor.get()<a name="line.290"></a>
<span class="sourceLineNo">291</span> .createStatus("Splitting " + wal + " to temporary staging area.", false, true);<a name="line.291"></a>
<span class="sourceLineNo">292</span> Reader walReader = null;<a name="line.292"></a>
<span class="sourceLineNo">293</span> this.fileBeingSplit = walStatus;<a name="line.293"></a>
<span class="sourceLineNo">294</span> long startTS = EnvironmentEdgeManager.currentTime();<a name="line.294"></a>
<span class="sourceLineNo">295</span> long length = walStatus.getLen();<a name="line.295"></a>
<span class="sourceLineNo">296</span> String lengthStr = StringUtils.humanSize(length);<a name="line.296"></a>
<span class="sourceLineNo">297</span> createOutputSinkAndEntryBuffers();<a name="line.297"></a>
<span class="sourceLineNo">298</span> try {<a name="line.298"></a>
<span class="sourceLineNo">299</span> String logStr = "Splitting " + wal + ", size=" + lengthStr + " (" + length + "bytes)";<a name="line.299"></a>
<span class="sourceLineNo">300</span> LOG.info(logStr);<a name="line.300"></a>
<span class="sourceLineNo">301</span> status.setStatus(logStr);<a name="line.301"></a>
<span class="sourceLineNo">302</span> if (cancel != null &amp;&amp; !cancel.progress()) {<a name="line.302"></a>
<span class="sourceLineNo">303</span> cancelled = true;<a name="line.303"></a>
<span class="sourceLineNo">304</span> return new SplitWALResult(false, corrupt);<a name="line.304"></a>
<span class="sourceLineNo">305</span> }<a name="line.305"></a>
<span class="sourceLineNo">306</span> walReader = getReader(walStatus, this.skipErrors, cancel);<a name="line.306"></a>
<span class="sourceLineNo">307</span> if (walReader == null) {<a name="line.307"></a>
<span class="sourceLineNo">308</span> LOG.warn("Nothing in {}; empty?", wal);<a name="line.308"></a>
<span class="sourceLineNo">309</span> return new SplitWALResult(true, corrupt);<a name="line.309"></a>
<span class="sourceLineNo">310</span> }<a name="line.310"></a>
<span class="sourceLineNo">311</span> LOG.info("Open {} took {}ms", wal, EnvironmentEdgeManager.currentTime() - startTS);<a name="line.311"></a>
<span class="sourceLineNo">312</span> int numOpenedFilesBeforeReporting = conf.getInt("hbase.splitlog.report.openedfiles", 3);<a name="line.312"></a>
<span class="sourceLineNo">313</span> int numOpenedFilesLastCheck = 0;<a name="line.313"></a>
<span class="sourceLineNo">314</span> outputSink.setReporter(cancel);<a name="line.314"></a>
<span class="sourceLineNo">315</span> outputSink.setStatus(status);<a name="line.315"></a>
<span class="sourceLineNo">316</span> outputSink.startWriterThreads();<a name="line.316"></a>
<span class="sourceLineNo">317</span> outputSinkStarted = true;<a name="line.317"></a>
<span class="sourceLineNo">318</span> Entry entry;<a name="line.318"></a>
<span class="sourceLineNo">319</span> startTS = EnvironmentEdgeManager.currentTime();<a name="line.319"></a>
<span class="sourceLineNo">320</span> while ((entry = getNextLogLine(walReader, wal, this.skipErrors)) != null) {<a name="line.320"></a>
<span class="sourceLineNo">321</span> byte[] region = entry.getKey().getEncodedRegionName();<a name="line.321"></a>
<span class="sourceLineNo">322</span> String encodedRegionNameAsStr = Bytes.toString(region);<a name="line.322"></a>
<span class="sourceLineNo">323</span> Long lastFlushedSequenceId = lastFlushedSequenceIds.get(encodedRegionNameAsStr);<a name="line.323"></a>
<span class="sourceLineNo">324</span> if (lastFlushedSequenceId == null) {<a name="line.324"></a>
<span class="sourceLineNo">325</span> if (<a name="line.325"></a>
<span class="sourceLineNo">326</span> !(isRegionDirPresentUnderRoot(entry.getKey().getTableName(), encodedRegionNameAsStr))<a name="line.326"></a>
<span class="sourceLineNo">327</span> ) {<a name="line.327"></a>
<span class="sourceLineNo">328</span> // The region directory itself is not present in the FS. This indicates that<a name="line.328"></a>
<span class="sourceLineNo">329</span> // the region/table is already removed. We can just skip all the edits for this<a name="line.329"></a>
<span class="sourceLineNo">330</span> // region. Setting lastFlushedSequenceId as Long.MAX_VALUE so that all edits<a name="line.330"></a>
<span class="sourceLineNo">331</span> // will get skipped by the seqId check below.<a name="line.331"></a>
<span class="sourceLineNo">332</span> // See more details at https://issues.apache.org/jira/browse/HBASE-24189<a name="line.332"></a>
<span class="sourceLineNo">333</span> LOG.info("{} no longer in filesystem; skipping all edits.", encodedRegionNameAsStr);<a name="line.333"></a>
<span class="sourceLineNo">334</span> lastFlushedSequenceId = Long.MAX_VALUE;<a name="line.334"></a>
<span class="sourceLineNo">335</span> } else {<a name="line.335"></a>
<span class="sourceLineNo">336</span> if (sequenceIdChecker != null) {<a name="line.336"></a>
<span class="sourceLineNo">337</span> RegionStoreSequenceIds ids = sequenceIdChecker.getLastSequenceId(region);<a name="line.337"></a>
<span class="sourceLineNo">338</span> Map&lt;byte[], Long&gt; maxSeqIdInStores = new TreeMap&lt;&gt;(Bytes.BYTES_COMPARATOR);<a name="line.338"></a>
<span class="sourceLineNo">339</span> for (StoreSequenceId storeSeqId : ids.getStoreSequenceIdList()) {<a name="line.339"></a>
<span class="sourceLineNo">340</span> maxSeqIdInStores.put(storeSeqId.getFamilyName().toByteArray(),<a name="line.340"></a>
<span class="sourceLineNo">341</span> storeSeqId.getSequenceId());<a name="line.341"></a>
<span class="sourceLineNo">342</span> }<a name="line.342"></a>
<span class="sourceLineNo">343</span> regionMaxSeqIdInStores.put(encodedRegionNameAsStr, maxSeqIdInStores);<a name="line.343"></a>
<span class="sourceLineNo">344</span> lastFlushedSequenceId = ids.getLastFlushedSequenceId();<a name="line.344"></a>
<span class="sourceLineNo">345</span> if (LOG.isDebugEnabled()) {<a name="line.345"></a>
<span class="sourceLineNo">346</span> LOG.debug("Last flushed sequenceid for " + encodedRegionNameAsStr + ": "<a name="line.346"></a>
<span class="sourceLineNo">347</span> + TextFormat.shortDebugString(ids));<a name="line.347"></a>
<span class="sourceLineNo">348</span> }<a name="line.348"></a>
<span class="sourceLineNo">349</span> }<a name="line.349"></a>
<span class="sourceLineNo">350</span> if (lastFlushedSequenceId == null) {<a name="line.350"></a>
<span class="sourceLineNo">351</span> lastFlushedSequenceId = -1L;<a name="line.351"></a>
<span class="sourceLineNo">352</span> }<a name="line.352"></a>
<span class="sourceLineNo">353</span> }<a name="line.353"></a>
<span class="sourceLineNo">354</span> lastFlushedSequenceIds.put(encodedRegionNameAsStr, lastFlushedSequenceId);<a name="line.354"></a>
<span class="sourceLineNo">355</span> }<a name="line.355"></a>
<span class="sourceLineNo">356</span> editsCount++;<a name="line.356"></a>
<span class="sourceLineNo">357</span> if (lastFlushedSequenceId &gt;= entry.getKey().getSequenceId()) {<a name="line.357"></a>
<span class="sourceLineNo">358</span> editsSkipped++;<a name="line.358"></a>
<span class="sourceLineNo">359</span> continue;<a name="line.359"></a>
<span class="sourceLineNo">360</span> }<a name="line.360"></a>
<span class="sourceLineNo">361</span> // Don't send Compaction/Close/Open region events to recovered edit type sinks.<a name="line.361"></a>
<span class="sourceLineNo">362</span> if (entry.getEdit().isMetaEdit() &amp;&amp; !outputSink.keepRegionEvent(entry)) {<a name="line.362"></a>
<span class="sourceLineNo">363</span> editsSkipped++;<a name="line.363"></a>
<span class="sourceLineNo">364</span> continue;<a name="line.364"></a>
<span class="sourceLineNo">365</span> }<a name="line.365"></a>
<span class="sourceLineNo">366</span> entryBuffers.appendEntry(entry);<a name="line.366"></a>
<span class="sourceLineNo">367</span> int moreWritersFromLastCheck = this.getNumOpenWriters() - numOpenedFilesLastCheck;<a name="line.367"></a>
<span class="sourceLineNo">368</span> // If sufficient edits have passed, check if we should report progress.<a name="line.368"></a>
<span class="sourceLineNo">369</span> if (<a name="line.369"></a>
<span class="sourceLineNo">370</span> editsCount % interval == 0 || moreWritersFromLastCheck &gt; numOpenedFilesBeforeReporting<a name="line.370"></a>
<span class="sourceLineNo">371</span> ) {<a name="line.371"></a>
<span class="sourceLineNo">372</span> numOpenedFilesLastCheck = this.getNumOpenWriters();<a name="line.372"></a>
<span class="sourceLineNo">373</span> String countsStr = (editsCount - (editsSkipped + outputSink.getTotalSkippedEdits()))<a name="line.373"></a>
<span class="sourceLineNo">374</span> + " edits, skipped " + editsSkipped + " edits.";<a name="line.374"></a>
<span class="sourceLineNo">375</span> status.setStatus("Split " + countsStr);<a name="line.375"></a>
<span class="sourceLineNo">376</span> if (cancel != null &amp;&amp; !cancel.progress()) {<a name="line.376"></a>
<span class="sourceLineNo">377</span> cancelled = true;<a name="line.377"></a>
<span class="sourceLineNo">378</span> return new SplitWALResult(false, corrupt);<a name="line.378"></a>
<span class="sourceLineNo">379</span> }<a name="line.379"></a>
<span class="sourceLineNo">380</span> }<a name="line.380"></a>
<span class="sourceLineNo">381</span> }<a name="line.381"></a>
<span class="sourceLineNo">382</span> } catch (InterruptedException ie) {<a name="line.382"></a>
<span class="sourceLineNo">383</span> IOException iie = new InterruptedIOException();<a name="line.383"></a>
<span class="sourceLineNo">384</span> iie.initCause(ie);<a name="line.384"></a>
<span class="sourceLineNo">385</span> throw iie;<a name="line.385"></a>
<span class="sourceLineNo">386</span> } catch (CorruptedLogFileException e) {<a name="line.386"></a>
<span class="sourceLineNo">387</span> LOG.warn("Could not parse, corrupt WAL={}", wal, e);<a name="line.387"></a>
<span class="sourceLineNo">388</span> // If splitLogWorkerCoordination, then its old-school zk-coordinated splitting so update<a name="line.388"></a>
<span class="sourceLineNo">389</span> // zk. Otherwise, it is the newer procedure-based WAL split which has no zk component.<a name="line.389"></a>
<span class="sourceLineNo">390</span> if (this.splitLogWorkerCoordination != null) {<a name="line.390"></a>
<span class="sourceLineNo">391</span> // Some tests pass in a csm of null.<a name="line.391"></a>
<span class="sourceLineNo">392</span> splitLogWorkerCoordination.markCorrupted(walRootDir, wal.getName(), walFS);<a name="line.392"></a>
<span class="sourceLineNo">393</span> }<a name="line.393"></a>
<span class="sourceLineNo">394</span> corrupt = true;<a name="line.394"></a>
<span class="sourceLineNo">395</span> } catch (IOException e) {<a name="line.395"></a>
<span class="sourceLineNo">396</span> e = e instanceof RemoteException ? ((RemoteException) e).unwrapRemoteException() : e;<a name="line.396"></a>
<span class="sourceLineNo">397</span> throw e;<a name="line.397"></a>
<span class="sourceLineNo">398</span> } finally {<a name="line.398"></a>
<span class="sourceLineNo">399</span> final String log = "Finishing writing output for " + wal + " so closing down";<a name="line.399"></a>
<span class="sourceLineNo">400</span> LOG.debug(log);<a name="line.400"></a>
<span class="sourceLineNo">401</span> status.setStatus(log);<a name="line.401"></a>
<span class="sourceLineNo">402</span> try {<a name="line.402"></a>
<span class="sourceLineNo">403</span> if (null != walReader) {<a name="line.403"></a>
<span class="sourceLineNo">404</span> walReader.close();<a name="line.404"></a>
<span class="sourceLineNo">405</span> }<a name="line.405"></a>
<span class="sourceLineNo">406</span> } catch (IOException exception) {<a name="line.406"></a>
<span class="sourceLineNo">407</span> LOG.warn("Could not close {} reader", wal, exception);<a name="line.407"></a>
<span class="sourceLineNo">408</span> }<a name="line.408"></a>
<span class="sourceLineNo">409</span> try {<a name="line.409"></a>
<span class="sourceLineNo">410</span> if (outputSinkStarted) {<a name="line.410"></a>
<span class="sourceLineNo">411</span> // Set cancelled to true as the immediate following statement will reset its value.<a name="line.411"></a>
<span class="sourceLineNo">412</span> // If close() throws an exception, cancelled will have the right value<a name="line.412"></a>
<span class="sourceLineNo">413</span> cancelled = true;<a name="line.413"></a>
<span class="sourceLineNo">414</span> cancelled = outputSink.close() == null;<a name="line.414"></a>
<span class="sourceLineNo">415</span> }<a name="line.415"></a>
<span class="sourceLineNo">416</span> } finally {<a name="line.416"></a>
<span class="sourceLineNo">417</span> long processCost = EnvironmentEdgeManager.currentTime() - startTS;<a name="line.417"></a>
<span class="sourceLineNo">418</span> // See if length got updated post lease recovery<a name="line.418"></a>
<span class="sourceLineNo">419</span> String msg = "Processed " + editsCount + " edits across "<a name="line.419"></a>
<span class="sourceLineNo">420</span> + outputSink.getNumberOfRecoveredRegions() + " Regions in " + processCost<a name="line.420"></a>
<span class="sourceLineNo">421</span> + " ms; skipped=" + editsSkipped + "; WAL=" + wal + ", size=" + lengthStr + ", length="<a name="line.421"></a>
<span class="sourceLineNo">422</span> + length + ", corrupted=" + corrupt + ", cancelled=" + cancelled;<a name="line.422"></a>
<span class="sourceLineNo">423</span> LOG.info(msg);<a name="line.423"></a>
<span class="sourceLineNo">424</span> status.markComplete(msg);<a name="line.424"></a>
<span class="sourceLineNo">425</span> if (LOG.isDebugEnabled()) {<a name="line.425"></a>
<span class="sourceLineNo">426</span> LOG.debug("Completed split of {}, journal: {}", wal, status.prettyPrintJournal());<a name="line.426"></a>
<span class="sourceLineNo">427</span> }<a name="line.427"></a>
<span class="sourceLineNo">428</span> }<a name="line.428"></a>
<span class="sourceLineNo">429</span> }<a name="line.429"></a>
<span class="sourceLineNo">430</span> return new SplitWALResult(!cancelled, corrupt);<a name="line.430"></a>
<span class="sourceLineNo">431</span> }<a name="line.431"></a>
<span class="sourceLineNo">432</span><a name="line.432"></a>
<span class="sourceLineNo">433</span> private boolean isRegionDirPresentUnderRoot(TableName tn, String region) throws IOException {<a name="line.433"></a>
<span class="sourceLineNo">434</span> return this.rootFS.exists(CommonFSUtils.getRegionDir(this.rootDir, tn, region));<a name="line.434"></a>
<span class="sourceLineNo">435</span> }<a name="line.435"></a>
<span class="sourceLineNo">436</span><a name="line.436"></a>
<span class="sourceLineNo">437</span> /**<a name="line.437"></a>
<span class="sourceLineNo">438</span> * Create a new {@link Reader} for reading logs to split.<a name="line.438"></a>
<span class="sourceLineNo">439</span> * @return Returns null if file has length zero or file can't be found.<a name="line.439"></a>
<span class="sourceLineNo">440</span> */<a name="line.440"></a>
<span class="sourceLineNo">441</span> protected Reader getReader(FileStatus walStatus, boolean skipErrors,<a name="line.441"></a>
<span class="sourceLineNo">442</span> CancelableProgressable cancel) throws IOException, CorruptedLogFileException {<a name="line.442"></a>
<span class="sourceLineNo">443</span> Path path = walStatus.getPath();<a name="line.443"></a>
<span class="sourceLineNo">444</span> long length = walStatus.getLen();<a name="line.444"></a>
<span class="sourceLineNo">445</span> Reader in;<a name="line.445"></a>
<span class="sourceLineNo">446</span><a name="line.446"></a>
<span class="sourceLineNo">447</span> // Check for possibly empty file. With appends, currently Hadoop reports a<a name="line.447"></a>
<span class="sourceLineNo">448</span> // zero length even if the file has been sync'd. Revisit if HDFS-376 or<a name="line.448"></a>
<span class="sourceLineNo">449</span> // HDFS-878 is committed.<a name="line.449"></a>
<span class="sourceLineNo">450</span> if (length &lt;= 0) {<a name="line.450"></a>
<span class="sourceLineNo">451</span> LOG.warn("File {} might be still open, length is 0", path);<a name="line.451"></a>
<span class="sourceLineNo">452</span> }<a name="line.452"></a>
<span class="sourceLineNo">453</span><a name="line.453"></a>
<span class="sourceLineNo">454</span> try {<a name="line.454"></a>
<span class="sourceLineNo">455</span> RecoverLeaseFSUtils.recoverFileLease(walFS, path, conf, cancel);<a name="line.455"></a>
<span class="sourceLineNo">456</span> try {<a name="line.456"></a>
<span class="sourceLineNo">457</span> in = getReader(path, cancel);<a name="line.457"></a>
<span class="sourceLineNo">458</span> } catch (EOFException e) {<a name="line.458"></a>
<span class="sourceLineNo">459</span> if (length &lt;= 0) {<a name="line.459"></a>
<span class="sourceLineNo">460</span> // TODO should we ignore an empty, not-last log file if skip.errors<a name="line.460"></a>
<span class="sourceLineNo">461</span> // is false? Either way, the caller should decide what to do. E.g.<a name="line.461"></a>
<span class="sourceLineNo">462</span> // ignore if this is the last log in sequence.<a name="line.462"></a>
<span class="sourceLineNo">463</span> // TODO is this scenario still possible if the log has been<a name="line.463"></a>
<span class="sourceLineNo">464</span> // recovered (i.e. closed)<a name="line.464"></a>
<span class="sourceLineNo">465</span> LOG.warn("Could not open {} for reading. File is empty", path, e);<a name="line.465"></a>
<span class="sourceLineNo">466</span> }<a name="line.466"></a>
<span class="sourceLineNo">467</span> // EOFException being ignored<a name="line.467"></a>
<span class="sourceLineNo">468</span> return null;<a name="line.468"></a>
<span class="sourceLineNo">469</span> }<a name="line.469"></a>
<span class="sourceLineNo">470</span> } catch (IOException e) {<a name="line.470"></a>
<span class="sourceLineNo">471</span> if (e instanceof FileNotFoundException) {<a name="line.471"></a>
<span class="sourceLineNo">472</span> // A wal file may not exist anymore. Nothing can be recovered so move on<a name="line.472"></a>
<span class="sourceLineNo">473</span> LOG.warn("File {} does not exist anymore", path, e);<a name="line.473"></a>
<span class="sourceLineNo">474</span> return null;<a name="line.474"></a>
<span class="sourceLineNo">475</span> }<a name="line.475"></a>
<span class="sourceLineNo">476</span> if (!skipErrors || e instanceof InterruptedIOException) {<a name="line.476"></a>
<span class="sourceLineNo">477</span> throw e; // Don't mark the file corrupted if interrupted, or not skipErrors<a name="line.477"></a>
<span class="sourceLineNo">478</span> }<a name="line.478"></a>
<span class="sourceLineNo">479</span> throw new CorruptedLogFileException("skipErrors=true; could not open " + path + ", skipping",<a name="line.479"></a>
<span class="sourceLineNo">480</span> e);<a name="line.480"></a>
<span class="sourceLineNo">481</span> }<a name="line.481"></a>
<span class="sourceLineNo">482</span> return in;<a name="line.482"></a>
<span class="sourceLineNo">483</span> }<a name="line.483"></a>
<span class="sourceLineNo">484</span><a name="line.484"></a>
<span class="sourceLineNo">485</span> private Entry getNextLogLine(Reader in, Path path, boolean skipErrors)<a name="line.485"></a>
<span class="sourceLineNo">486</span> throws CorruptedLogFileException, IOException {<a name="line.486"></a>
<span class="sourceLineNo">487</span> try {<a name="line.487"></a>
<span class="sourceLineNo">488</span> return in.next();<a name="line.488"></a>
<span class="sourceLineNo">489</span> } catch (EOFException eof) {<a name="line.489"></a>
<span class="sourceLineNo">490</span> // truncated files are expected if a RS crashes (see HBASE-2643)<a name="line.490"></a>
<span class="sourceLineNo">491</span> LOG.info("EOF from {}; continuing.", path);<a name="line.491"></a>
<span class="sourceLineNo">492</span> return null;<a name="line.492"></a>
<span class="sourceLineNo">493</span> } catch (IOException e) {<a name="line.493"></a>
<span class="sourceLineNo">494</span> // If the IOE resulted from bad file format,<a name="line.494"></a>
<span class="sourceLineNo">495</span> // then this problem is idempotent and retrying won't help<a name="line.495"></a>
<span class="sourceLineNo">496</span> if (<a name="line.496"></a>
<span class="sourceLineNo">497</span> e.getCause() != null &amp;&amp; (e.getCause() instanceof ParseException<a name="line.497"></a>
<span class="sourceLineNo">498</span> || e.getCause() instanceof org.apache.hadoop.fs.ChecksumException)<a name="line.498"></a>
<span class="sourceLineNo">499</span> ) {<a name="line.499"></a>
<span class="sourceLineNo">500</span> LOG.warn("Parse exception from {}; continuing", path, e);<a name="line.500"></a>
<span class="sourceLineNo">501</span> return null;<a name="line.501"></a>
<span class="sourceLineNo">502</span> }<a name="line.502"></a>
<span class="sourceLineNo">503</span> if (!skipErrors) {<a name="line.503"></a>
<span class="sourceLineNo">504</span> throw e;<a name="line.504"></a>
<span class="sourceLineNo">505</span> }<a name="line.505"></a>
<span class="sourceLineNo">506</span> throw new CorruptedLogFileException("skipErrors=true Ignoring exception"<a name="line.506"></a>
<span class="sourceLineNo">507</span> + " while parsing wal " + path + ". Marking as corrupted", e);<a name="line.507"></a>
<span class="sourceLineNo">508</span> }<a name="line.508"></a>
<span class="sourceLineNo">509</span> }<a name="line.509"></a>
<span class="sourceLineNo">510</span><a name="line.510"></a>
<span class="sourceLineNo">511</span> /**<a name="line.511"></a>
<span class="sourceLineNo">512</span> * Create a new {@link WALProvider.Writer} for writing log splits.<a name="line.512"></a>
<span class="sourceLineNo">513</span> * @return a new Writer instance, caller should close<a name="line.513"></a>
<span class="sourceLineNo">514</span> */<a name="line.514"></a>
<span class="sourceLineNo">515</span> protected WALProvider.Writer createWriter(Path logfile) throws IOException {<a name="line.515"></a>
<span class="sourceLineNo">516</span> return walFactory.createRecoveredEditsWriter(walFS, logfile);<a name="line.516"></a>
<span class="sourceLineNo">517</span> }<a name="line.517"></a>
<span class="sourceLineNo">518</span><a name="line.518"></a>
<span class="sourceLineNo">519</span> /**<a name="line.519"></a>
<span class="sourceLineNo">520</span> * Create a new {@link Reader} for reading logs to split.<a name="line.520"></a>
<span class="sourceLineNo">521</span> * @return new Reader instance, caller should close<a name="line.521"></a>
<span class="sourceLineNo">522</span> */<a name="line.522"></a>
<span class="sourceLineNo">523</span> private Reader getReader(Path curLogFile, CancelableProgressable reporter) throws IOException {<a name="line.523"></a>
<span class="sourceLineNo">524</span> return walFactory.createReader(walFS, curLogFile, reporter);<a name="line.524"></a>
<span class="sourceLineNo">525</span> }<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> * Get current open writers<a name="line.528"></a>
<span class="sourceLineNo">529</span> */<a name="line.529"></a>
<span class="sourceLineNo">530</span> private int getNumOpenWriters() {<a name="line.530"></a>
<span class="sourceLineNo">531</span> int result = 0;<a name="line.531"></a>
<span class="sourceLineNo">532</span> if (this.outputSink != null) {<a name="line.532"></a>
<span class="sourceLineNo">533</span> result += this.outputSink.getNumOpenWriters();<a name="line.533"></a>
<span class="sourceLineNo">534</span> }<a name="line.534"></a>
<span class="sourceLineNo">535</span> return result;<a name="line.535"></a>
<span class="sourceLineNo">536</span> }<a name="line.536"></a>
<span class="sourceLineNo">537</span><a name="line.537"></a>
<span class="sourceLineNo">538</span> /**<a name="line.538"></a>
<span class="sourceLineNo">539</span> * Contains some methods to control WAL-entries producer / consumer interactions<a name="line.539"></a>
<span class="sourceLineNo">540</span> */<a name="line.540"></a>
<span class="sourceLineNo">541</span> public static class PipelineController {<a name="line.541"></a>
<span class="sourceLineNo">542</span> // If an exception is thrown by one of the other threads, it will be<a name="line.542"></a>
<span class="sourceLineNo">543</span> // stored here.<a name="line.543"></a>
<span class="sourceLineNo">544</span> AtomicReference&lt;Throwable&gt; thrown = new AtomicReference&lt;&gt;();<a name="line.544"></a>
<span class="sourceLineNo">545</span><a name="line.545"></a>
<span class="sourceLineNo">546</span> // Wait/notify for when data has been produced by the writer thread,<a name="line.546"></a>
<span class="sourceLineNo">547</span> // consumed by the reader thread, or an exception occurred<a name="line.547"></a>
<span class="sourceLineNo">548</span> final Object dataAvailable = new Object();<a name="line.548"></a>
<span class="sourceLineNo">549</span><a name="line.549"></a>
<span class="sourceLineNo">550</span> void writerThreadError(Throwable t) {<a name="line.550"></a>
<span class="sourceLineNo">551</span> thrown.compareAndSet(null, t);<a name="line.551"></a>
<span class="sourceLineNo">552</span> }<a name="line.552"></a>
<span class="sourceLineNo">553</span><a name="line.553"></a>
<span class="sourceLineNo">554</span> /**<a name="line.554"></a>
<span class="sourceLineNo">555</span> * Check for errors in the writer threads. If any is found, rethrow it.<a name="line.555"></a>
<span class="sourceLineNo">556</span> */<a name="line.556"></a>
<span class="sourceLineNo">557</span> void checkForErrors() throws IOException {<a name="line.557"></a>
<span class="sourceLineNo">558</span> Throwable thrown = this.thrown.get();<a name="line.558"></a>
<span class="sourceLineNo">559</span> if (thrown == null) {<a name="line.559"></a>
<span class="sourceLineNo">560</span> return;<a name="line.560"></a>
<span class="sourceLineNo">561</span> }<a name="line.561"></a>
<span class="sourceLineNo">562</span> if (thrown instanceof IOException) {<a name="line.562"></a>
<span class="sourceLineNo">563</span> throw new IOException(thrown);<a name="line.563"></a>
<span class="sourceLineNo">564</span> } else {<a name="line.564"></a>
<span class="sourceLineNo">565</span> throw new RuntimeException(thrown);<a name="line.565"></a>
<span class="sourceLineNo">566</span> }<a name="line.566"></a>
<span class="sourceLineNo">567</span> }<a name="line.567"></a>
<span class="sourceLineNo">568</span> }<a name="line.568"></a>
<span class="sourceLineNo">569</span><a name="line.569"></a>
<span class="sourceLineNo">570</span> static class CorruptedLogFileException extends Exception {<a name="line.570"></a>
<span class="sourceLineNo">571</span> private static final long serialVersionUID = 1L;<a name="line.571"></a>
<span class="sourceLineNo">572</span><a name="line.572"></a>
<span class="sourceLineNo">573</span> CorruptedLogFileException(String s) {<a name="line.573"></a>
<span class="sourceLineNo">574</span> super(s);<a name="line.574"></a>
<span class="sourceLineNo">575</span> }<a name="line.575"></a>
<span class="sourceLineNo">576</span><a name="line.576"></a>
<span class="sourceLineNo">577</span> /**<a name="line.577"></a>
<span class="sourceLineNo">578</span> * CorruptedLogFileException with cause<a name="line.578"></a>
<span class="sourceLineNo">579</span> * @param message the message for this exception<a name="line.579"></a>
<span class="sourceLineNo">580</span> * @param cause the cause for this exception<a name="line.580"></a>
<span class="sourceLineNo">581</span> */<a name="line.581"></a>
<span class="sourceLineNo">582</span> CorruptedLogFileException(String message, Throwable cause) {<a name="line.582"></a>
<span class="sourceLineNo">583</span> super(message, cause);<a name="line.583"></a>
<span class="sourceLineNo">584</span> }<a name="line.584"></a>
<span class="sourceLineNo">585</span> }<a name="line.585"></a>
<span class="sourceLineNo">586</span>}<a name="line.586"></a>
</pre>
</div>
</body>
</html>