Cloudera Manager Agents
The Cloudera Manager Agent is a Cloudera Manager component that works with the Cloudera Manager Server to manage the processes that map to role instances.
In a Cloudera Manager managed cluster, you can only start or stop role instance processes
using Cloudera Manager. Cloudera Manager uses a modified version of the open source process
management tool, Supervisor. The Cloudera Manager Agent manages this through a process called
cloudera-scm-supervisord, which starts processes, redirects log files,
notifies of process failures, and sets the effective user ID for service processes. Cloudera
Manager supports automatically restarting a crashed process. It will also flag a role instance
with a bad health flag if its process crashes repeatedly right after start up.
The Agent is started by init.d at
start-up. It, in turn, contacts the Cloudera Manager Server and determines
which processes should be running. The Agent is monitored as part of
Cloudera Manager's host monitoring. If the Agent stops heartbeating, the
host is marked as having bad health.
One of the Agent's main responsibilities is to start and stop processes.
When the Agent detects a new process from the Server heartbeat, the Agent creates a directory
for it in /var/run/cloudera-scm-agent and unpacks the configuration. It then
contacts cloudera-scm-supervisord, which starts the process.
Monitored File Systems
ext2ext3ext4xfsaufstmpfs
By modifying the parameter monitored_nodev_filesystem_types in the
Cloudera Manager Agent's config.ini file, you can also add NFS type
mounts to the monitored list. For more information on this parameter, see Configuring Cloudera Manager Agents.
Other file system types, especially auto-mounted file systems, or other network file systems, are not supported and might prevent the operation of the Cloudera Manager Agent.
cm_processes
To enable Cloudera Manager to run scripts in subdirectories of
/var/run/cloudera-scm-agent, (because /var/run is
mounted noexec in many Linux distributions), Cloudera Manager mounts a
tmpfs, named cm_processes, for process subdirectories.
A tmpfs defaults to a max size of 50% of physical RAM
but this space is not allocated until its used, and tmpfs is paged out
to swap if there is memory pressure.
The lifecycle actions of cmprocesses can be described
by the following statements:
- Created when the Agent starts up for the first time with a new
cloudera-scm-supervisordprocess. - If it already exists without
noexec, reused when the Agent is started usingstartand not recreated. - Remounted if Agent is started using
clean_restart. - Unmounting and remounting cleans out the contents (since it is mounted as a tmpfs).
- Unmounted when the host is rebooted.
- Not unmounted when the Agent is stopped.
