ecclusterupgrade Options

Option Description

--admin-password=value

Cluster Manager administrator’s password. Default is changeme.

--agentinstalldir=path

Agent side install directory. You must specify the --hosts option also to limit the upgrade to those machines that use the “non-standard” agent install directory.

--batchsize=int

Number of hosts to process per iteration. This lets you break the upgrade of a large cluster into more manageable smaller batches to avoid timeouts. Default is 20 (the recommended maximum).

--cm=host[:port[:securePort]]

Cluster Manager whose hosts are to be upgraded. Default is localhost:8030:8031.

--cm-secure-port=port

Cluster Manager secure port. Default is 8031.

`--gen-hostlist `

Query the Cluster Manager to get the list of hosts, write it to a file, and exit.

`--help `

Display this table’s help text.

--hosts="host1 host2 …​"

List of hosts to upgrade. Can be a pattern such as host[1-3], host[2,5]. This switch supersedes --cm.

--installer=filename

Installer executable to use. Use an absolute filename for this option. If the installer is not in the expected directory, you must provide the full path to the installer, even when you specify --installerdir.

--installerdir=dirname

Directory containing the installer executables.

--installerversion=version

Installer version (for example, 7.0.0.12345 ). This option is needed only if there are multiple installer versions in the `local/--installerdir ` location.

--installtimeout=value

Number of minutes to wait for the agent install to finish. Default is 20.

--logfile=value

Log file for messages. The default is /var/tmp (Linux and Solaris) or C:\ ` (Windows) in the form of `ecloud_node_upgrade_YYYMMDD_HHMMSS.log.

--newcm=hostname/ip

Host name or IP address of the new Cluster Manager where you want to migrate the upgraded agents.

--remotetmpdir=path

Location where the remote install is run.

--targetos=os

Agent machine operating system ( windows, linux, or solaris ).