Use the SA Core Recertification tool

To run the Core recertification tool, enter the following:

/opt/opsware/oi_util/OpswareCertTool/recert_utils/corerecert [--phase <phase number>] [--config <complete path to the config file>] [--doit]] [-h, --help] [-v, --version] [-s, --status] [-d, --debug] [--summary] [--cancel_all_agent_recert_jobs] [--cancel_agent_recert_jobs_for_facility <facility name>] [--cancel_all_jobs] [--reason <reason for job cancellation>] [--force_resume <facility_name>]

The following topics are discussed in this section:

• Arguments to the Core Recertification tool
• Security considerations

Arguments to the Core Recertification tool

The following table describes the valid arguments for the Core Recertification tool:

Core recertification tool arguments

Argument	Description
-h, --help	Displays help.
--phase	Starts a specified Core Recertification phase. The valid phase numbers are 1, 4, 6, 7, 8, 9, 12, and 13.
--config <config file>	The fully qualified path to the Core Recertification configuration file. The default configuration file is /opt/opsware/oi_util/OpswareCertTool/recert_utils/corerecert.conf.
--doit	Reruns or forces a rerun of a given Core Recertification phase. This is useful when certain newly added components have missed the recertification process. It is also used to skip specified phases, such as new Agent CA push or old Agent CA removal.
-v, --version	Prints out the version number of the corerecert executable.
-s, --status	Displays the current status of the recertification process.
-d, --debug	Sets Core Recertification to debug mode, debug logs are available in /tmp/recerttool.log.
--summary	Prints out the current status summary, shorter version of --status.
--cancel_all_agent_recert_jobs	Cancels all currently scheduled Agent recertification jobs.
--cancel_agent_recert_jobs_for_ facility <facility name>	Cancels the Agent recertification jobs scheduled for a given facility.
--cancel_all_jobs	Cancels all Core and Agent Recertification jobs.
--reason <reason for job cancellation>	Specifies an optional reason for the job cancellation.
--force_resume	Specifies that a new job be automatically scheduled for any facilities with failed agent recertification jobs. Facilities with no failed jobs will be skipped. Alternatively, if you do not specify this parameter, you can resume the job for each facility individually.

The /tmp/recerttool.log is not cumulative, it is rewritten with each recerttool execution. The log contains only the following information: information on starting the background processes for the current phase, parameters that the current phase uses (if applicable), and information on failure to kick off background jobs.

The core recertification background jobs rely on SA's OGSH infrastructure. See /tmp/core_recert.log (SA 9.1, 10.00 and 10.10) under OGFS of the core used to start the recertification or in /var/log/opsware/waybot/recert.log (SA 10.20 and later) for more information.

The agent recertification background jobs are run by the waybot, hence more details can be found in the twist and waybot logs on each core of the mesh.

The Software and Model Repository signature regeneration (Phase 11) will log additional information on the recert's base slice in UpdateSignatures.log and ResignJobTokens.log under /opt/opsware/oi_util/OpswareCertTool/recert_utils/ .

Adding new Core Components during Core Recertification is not recommended. Although adding new Core Components, such as the Slice Component bundle, a Satellite, etc. during Core Recertification is possible under certain circumstances, HPE does not recommend doing so unless absolutely necessary. You must first contact HPE Professional Services before adding new Core components while a Core Recertification is in progress.

Security considerations

Consider the following security issues:

Crypto database file

The SA Core Recertification Tool requires access to the SA crypto database file during recertification.

The SA crypto database consists of the file:

/var/opt/opsware/crypto/cadb/realm/opsware-crypto.db.e

This file is protected by the crypto material password (decrypt_passwd), which was specified during the mesh’s First Core installation. During subsequent Core installations, this file is also copied to the new Secondary Core hosts. You must protect this password as compromising the crypto database files means compromising your entire Multimaster Mesh.

The crypto database file is required only during SA installation or upgrade, but it is regenerated during Core Recertification. Therefore, HPE strongly recommends that you create procedures that protect the crypto database file. Therefore, before Core Recertification, you must back up this file to a secure location.

During Core Recertification, SA regenerates the crypto database only on the host on which you invoke the Core Recertification Tool. Core Recertification does not copy the newly generated crypto database file to any other hosts in the mesh during recertification. You should also back up this file to a secure location as soon as Core Recertification is complete.

Equally important is to strictly control root access to the Core hosts. Crypto materials (certificates and their corresponding private keys) on the Core hosts are not encrypted. They are protected by the root user account. In other words, these files are protected by the read-only access for the root user. Therefore, having root access to the Core hosts means a user has access to both the crypto material password and the crypto database files, and Core Recertification should only be performed by SA System Administrators, or someone who has legitimate root access to the Core hosts.