Github user tgravescs commented on a diff in the pull request:
https://github.com/apache/spark/pull/204#discussion_r11389264
--- Diff:
core/src/main/scala/org/apache/spark/deploy/history/HistoryServer.scala ---
@@ -0,0 +1,265 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.deploy.history
+
+import javax.servlet.http.HttpServletRequest
+
+import scala.collection.mutable
+import scala.concurrent._
+import scala.concurrent.ExecutionContext.Implicits.global
+
+import org.apache.hadoop.fs.{FileStatus, Path}
+import org.eclipse.jetty.servlet.ServletContextHandler
+
+import org.apache.spark.{Logging, SecurityManager, SparkConf}
+import org.apache.spark.deploy.SparkUIContainer
+import org.apache.spark.ui.SparkUI
+import org.apache.spark.ui.JettyUtils._
+import org.apache.spark.util.Utils
+import org.apache.spark.scheduler.{ApplicationEventListener,
ReplayListenerBus}
+
+/**
+ * A web server that renders SparkUIs of finished applications.
+ *
+ * For the standalone mode, MasterWebUI already achieves this
functionality. Thus, the
+ * main use case of the HistoryServer is in other deploy modes (e.g. Yarn
or Mesos).
+ *
+ * The logging directory structure is as follows: Within the given base
directory, each
+ * application's event logs are maintained in the application's own
sub-directory. This
+ * is the same structure as maintained in the event log write code path in
+ * EventLoggingListener.
+ *
+ * @param baseLogDir The base directory in which event logs are found
+ * @param requestedPort The requested port to which this server is to be
bound
+ */
+class HistoryServer(
+ val baseLogDir: String,
+ requestedPort: Int,
+ conf: SparkConf)
+ extends SparkUIContainer("History Server") with Logging {
+
+ import HistoryServer._
+
+ private val fileSystem = Utils.getHadoopFileSystem(baseLogDir)
+ private val bindHost = Utils.localHostName()
+ private val publicHost =
Option(System.getenv("SPARK_PUBLIC_DNS")).getOrElse(bindHost)
+ private val port = requestedPort
+ private val securityManager = new SecurityManager(conf)
+ private val indexPage = new IndexPage(this)
+
+ // A timestamp of when the disk was last accessed to check for log
updates
+ private var lastLogCheck = -1L
+
+ private val handlers = Seq[ServletContextHandler](
+ createStaticHandler(STATIC_RESOURCE_DIR, "/static"),
+ createServletHandler("/",
+ (request: HttpServletRequest) => indexPage.render(request),
securityMgr = securityManager)
+ )
+
+ // A mapping of application ID to its history information, which
includes the rendered UI
+ val appIdToInfo = mutable.HashMap[String, ApplicationHistoryInfo]()
+
+ // A set of recently removed applications that the server should avoid
re-rendering
+ val appIdBlacklist = mutable.HashSet[String]()
--- End diff --
This isn't ideal for scaling. At some point this is going to get huge
unless the application history logs are cleaned up. Perhaps we leave that for
a follow on improvement though since I think you are limited by the way its
storing them now. The first thing that came to mind was just to list by date
and choose the latest x but hadoop filesystem doesn't give you that filter
without doing stat on each one afterwards.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
