RegionServer Health Tests

This is a RegionServer health test that checks whether the Master considers the RegionServer alive. A failure of this health test may indicate that the RegionServer is having trouble communicating with at least the HBase Master and possibly the entire HBase cluster. Look in the RegionServer logs for more details. This test can be enabled or disabled using the HBase RegionServer to Master Connectivity Test RegionServer monitoring setting. The HBase Region Server Connectivity Tolerance at Startup RegionServer monitoring setting and the Health Check Startup Tolerance Master monitoring setting can be used to control the test's tolerance windows around RegionServer and Master restarts respectively.

Short Name: Cluster Connectivity

This is a RegionServer health test that checks that a moving average of the size of the RegionServer's compaction queue does not exceed some value. A failure of this health test indicates a high write load on the RegionServer. Try reducing the write load on the RegionServer, increasing capacity by adding additional disks to the RegionServer, or adding additional RegionServers. This test can be configured using the HBase RegionServer Compaction Queue Monitoring Thresholds and HBase RegionServer Compaction Queue Monitoring Period RegionServer monitoring settings.

Short Name: Compaction Queue Size

This RegionServer health test checks that the number of file descriptors used does not rise above some percentage of the RegionServer file descriptor limit. A failure of this health test may indicate a bug in either Hadoop or Cloudera Manager. Contact Cloudera support. This test can be configured using the File Descriptor Monitoring Thresholds RegionServer monitoring setting.

Short Name: File Descriptors

This is a RegionServer health test that checks that a moving average of the size of the RegionServer's flush queue does not exceed some value. A failure of this health test indicates a high write load on the RegionServer. Try reducing the write load on the RegionServer, increasing capacity by adding additional disks to the RegionServer, or adding additional RegionServers. This test can be configured using the HBase RegionServer Flush Queue Monitoring Thresholds and HBase RegionServer Flush Queue Monitoring Period RegionServer monitoring settings.

Short Name: Flush Queue Size

This RegionServer health test checks that the RegionServer is not spending too much time performing Java garbage collection. It checks that no more than some percentage of recent time is spent performing Java garbage collection. A failure of this health test may indicate a capacity planning problem or misconfiguration of the RegionServer. This test can be configured using the Garbage Collection Duration Thresholds and Garbage Collection Duration Monitoring Period RegionServer monitoring settings.

Short Name: GC Duration

This is a RegionServer health test that checks that the average time it takes for the RegionServer to read from HDFS does not exceed some value. The health test computes a moving average of the average HDFS read time for the RegionServer over a configurable window and then compares that moving average to a configured threshold value. A moving average is used to avoid alerting on short-lived HDFS read latency spikes. The behavior of this health test depends on how the HBase cluster is being used. A failure of this health test may indicate a problem with HDFS, extensive load on the cluster, or a capacity management problem. For example, it is possible to see average read latency increase while executing MapReduce jobs on the cluster. In such situations, if the read latency increase is unacceptable, consider running the MapReduce jobs with fewer map or reduce slots. This test can be configured using the HBase RegionServer HDFS Read Latency Thresholds and HBase RegionServer HDFS Read Latency Monitoring Period RegionServer monitoring settings.

Short Name: HDFS Read Latency

This is a RegionServer health test that checks that the average time it takes for the RegionServer to perform HDFS syncs does not exceed some value. The health test computes a moving average of the average HDFS sync time for the RegionServer over a configurable window and then compares that moving average to a configured threshold value. A moving average is used to avoid alerting on short-lived HDFS sync latency spikes. The behavior of this health test depends on how the HBase cluster is being used. A failure of this health test may indicate a problem with HDFS, extensive load on the cluster, or a capacity management problem. For example, it is possible to see average sync latency increase while executing MapReduce jobs on the cluster. In such situations, if the sync latency increase is unacceptable, consider running the MapReduce jobs with fewer map or reduce slots. This test can be configured using the HBase RegionServer HDFS Sync Latency Thresholds and HBase RegionServer HDFS Sync Latency Monitoring Period RegionServer monitoring settings.

Short Name: HDFS Sync Latency

This RegionServer health test checks that the filesystem containing the heap dump directory of this RegionServer has sufficient free space. This test can be configured using the Heap Dump Directory Free Space Monitoring Absolute Thresholds and Heap Dump Directory Free Space Monitoring Percentage Thresholds RegionServer monitoring settings.

Short Name: Heap Dump Directory Free Space

This RegionServer health test factors in the health of the host upon which the RegionServer is running. A failure of this test means that the host running the RegionServer is experiencing some problem. See that host's status page for more details.This test can be enabled or disabled using the RegionServer Host Health Test RegionServer monitoring setting.

Short Name: Host Health

This RegionServer health test checks that the filesystem containing the log directory of this RegionServer has sufficient free space. This test can be configured using the Log Directory Free Space Monitoring Absolute Thresholds and Log Directory Free Space Monitoring Percentage Thresholds RegionServer monitoring settings.

Short Name: Log Directory Free Space

This is a RegionServer health test that checks that the amount of the RegionServer's memory devoted to memstores does not exceed some percentage of the RegionServer's configured hbase.regionserver.global.memstore.upperLimit. When a RegionServer's memstores reach this maximum size, new updates are blocked while the RegionServer flushes. A failure of this health test indicates a high write load on the RegionServer. Try reducing the write load on the RegionServer, increasing capacity by adding additional disks to the RegionServer, or adding additional RegionServers. This test can be configured using the HBase RegionServer Memstore Size Thresholds RegionServer monitoring setting.

Short Name: Memstore Size

This RegionServer health test checks that the Cloudera Manager Agent on the RegionServer host is heart beating correctly and that the process associated with the RegionServer role is in the state expected by Cloudera Manager. A failure of this health test may indicate a problem with the RegionServer process, a lack of connectivity to the Cloudera Manager Agent on the RegionServer host, or a problem with the Cloudera Manager Agent. This test can fail either because the RegionServer has crashed or because the RegionServer will not start or stop in a timely fashion. Check the RegionServer logs for more details. If the test fails because of problems communicating with the Cloudera Manager Agent on the RegionServer host, check the status of the Cloudera Manager Agent by running /etc/init.d/cloudera-scm-agent status on the RegionServer host, or look in the Cloudera Manager Agent logs on the RegionServer host for more details. This test can be enabled or disabled using the RegionServer Process Health Test RegionServer monitoring setting.

Short Name: Process Status

This is a RegionServer health test that checks that the sum of the sizes of all store file indexes does not exceed some percentage of the RegionServer's maximum heap size. A failure of this health test indicates that the RegionServer is using a significant portion of its memory for store file indexes. If the amount of memory devoted to these indexes is undesirably high, the size of indexes can be reduced by increasing the HBase block size, by using smaller key values, or by using fewer columns. Each of these choices involves trade-offs. Contact Cloudera Support for more information on this topic. This test can be configured using the Percentage of Heap Used by HStoreFile Index RegionServer monitoring setting.

Short Name: Store File Index Size

This RegionServer health test checks the amount of swap memory in use by the role. A failure of this health test may indicate that your machine is overloaded. This test can be configured using the Process Swap Memory Thresholds monitoring settings.

Short Name: Swap Memory Usage

This RegionServer health test checks that the RegionServer has not recently exited unexpectedly. The test returns "Bad" health if the number of unexpected exits goes above a critical threshold. For example, if this test is configured with a critical threshold of 1, this test would return "Good" health if there have been no unexpected exits recently. If there has been 1 or more unexpected exits recently, this test would return "Bad" health. This test can be configured using the Unexpected Exits Thresholds and Unexpected Exits Monitoring Period RegionServer monitoring settings.

Short Name: Unexpected Exits

This health test checks that the role's web server is responding quickly to requests by the Cloudera Manager Agent, and that the Cloudera Manager Agent can collect metrics from the web server. Failure of this health test may indicate a problem with the web server of the RegionServer, a misconfiguration of the RegionServer, or a problem with the Cloudera Manager Agent. Consult the Cloudera Manager Agent logs and the logs of the RegionServer for more detail. If the test failure message indicates a communication problem, the Cloudera Manager Agent's HTTP requests to the RegionServer's web server are failing or timing out. If the test's failure message indicates an unexpected response, the RegionServer's web server responded to the Cloudera Manager Agent's request, but the response could not be interpreted for some reason. This test can be configured using the Web Metric Collection RegionServer monitoring setting.

Short Name: Web Server Status