Skip to content

basho/cluster_info

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

The cluster_info application

Erlang CI Actions Status

The cluster_info application provides a flexible and easily-extendible way to dump the state of a cluster of Erlang nodes.

Some of the information that the application can gather includes:

  • Date & time
  • Statistics on all Erlang processes on the node
  • Network connection details to all other Erlang nodes
  • Top CPU- and memory-hogging processes
  • Processes with large mailboxes
  • Internal memory allocator statistics
  • ETS table information
  • The names & versions of each code module loaded into the node

The app can also automatically gather all of this data from all nodes and write it into a single file. It's about as easy as can be to take a snapshot of all nodes in a cluster. It is a valuable tool for support and development teams to diagnose problems in a cluster, as a tool to aid capacity planning, and merely to answer a curious question like, "What's really going on in there?"

Example usage

  • cluster_info:dump_all_connected("/tmp/report.all-nodes.txt").
  • cluster_info:dump_local_node("/tmp/report.local-node.txt").
  • cluster_info:dump_nodes([riak@boxA, riak@boxB], "/tmp/report.some-nodes.txt").

See the file example-output.Riak.txt (in the top of the repository) for sample output from a single-node Riak system. Use the regular expression ^==* to find major & minor sections within the file. (NOTE The regular expression has a space character at the end of it.)

Configuring maximum string length formatting limits

It's a well-known feature of Erlang that the default string representation is a list of byte values. Strings can consume much more RAM than if the equivalent data were stored in the Erlang binary data type instead. If a prerequisite OTP application, lager, is available, then this application can use the lager_format:format/4 function to attempt to limit the amount of memory used while generating reports.

To limit the amount of RAM used to format strings in a report, this application will attempt to fetch the following OTP application environment variable from the cluster_info app's variables:

  • fmt_max_bytes When formatting a log-related term that might be "big", limit the term's formatted output to a maximum of fmt_max_bytes bytes.

The default value is 256KB.

You have several options for configuring these OTP application environment variables:

  • Add the following to your packaging's system configuration file (which is specified by the -config /path/to/file flag to the runtime system ... see the online docs for configuration OTP applications, "7.8 Configuring an Application" for more details:

    {cluster_info, [{term_max_size, 65536}, {fmt_max_bytes, 65536}]}

  • Add the following the env section of your copy of the package's ebin/cluster_info.app file

    {env, [{term_max_size, 65536}, {fmt_max_bytes, 65536}]}

  • Execute the following code to set the environment variables at runtime. Please note that setting these parameters will only affect processes that are created after the values are set:

    application:set_env(cluster_info, term_max_size, 65536), application:set_env(cluster_info, fmt_max_bytes, 65536)

If the fmt_max_bytes environment variable is not defined, then the length of formatted strings will not be restricted.

Licensing

The cluster_info application was written by Gemini Mobile Technologies, Inc. and is licensed under the Apache Public License version 2.0. This fork of the code has been subsequently modified by Basho Technologies, Inc. and is distributed under the same license.