Oracle SuperCluster M6-32 Hardware
 
Oracle SuperCluster T5-8 Hardware
 
Oracle Platinum Services
 
Oracle SuperCluster T5-8 Half Rack
 
Oracle SuperCluster Specific Software
 

PLA-Support>Eng Systems>Exadata/ODA/SSC>SPARC SuperCluster>DB: SuperCluster_EST
 
Tools>Primary Use>Data Collection
 

In this Document

Goal

Solution

Setting up Exachk in your SuperCluster Environment

Notes and considerations

Deployment

Merging collections

Advanced Setup

References

Applies to:

Goal

SuperCluster Best Practices for using Exachk

- Download and install Exachk

- Exachk in Global vs. Local Zones

- Exachk in Application vs. Database zones

- Merging Exachk data collections for upload

- Proactive monitoring with Exachk

Solution

Setting up Exachk in your SuperCluster Environment

Once installed and used correctly on a regular basis, Oracle's Engineered Systems Configuration Audit Tool (aka Exachk) utility can be an invaluable aid in keeping your system up to date and recording changes, etc. as well as providing baselines for support analysis in the event such data is required to resolve a Service Request. There are configuration management utilities available to manage collections from across all Engineered Systems in a given environment. 

When exachk is run from the primary ldom as user 'root' it will discover all RAC/database instances and compute nodes and all storage servers and InfiniBand switches in the fabric and run exachk utilities specific to each.

When it is run in a DB zone or virtualized environment it will collect data from one or all instances / RAC nodes.

Command line options are available to run exachk on subset of system components; e.g. -clusternodes to designate Database Servers, -cells to designate Storage Servers and -ibswitches to designate InfiniBand Switches, etc.

See Exachk User's Guide for Engineered Systems Configuration Audit Tool for complete command line details.

Notes and considerations

If installed and run as root It will collect connectivity data from the GZ to the storage cells and other nodes; however it will only collect DB/RAC data if run as user root from within nodes or zones in which RDMBS and / or RAC software is installed.

By default there should be no need to set environment variables or run as any other user than root however there may be exceptions when one is asked to run it in some debug or extra-ordinary fashion by an Oracle Support Analyst. 

There are projects underway to make Exachk run without requiring super user privileges and without having root ssh/RSH key equivalency setup; however, currently it is best practice to have things set up within a given SuperCluster to allow Exachk to run as currently implemented.

The output destination is /opt/oracle.SupportTools/exachk unless otherwise specified. Set environment variable RAT_OUTPUT to change this. E.g.

Multiple collections can be merged into one file for ease of uploading to MOS for analysis; e.g. data collections from multiple GZ and NGZ environments for a single case, see example.

Deployment

1. Download the latest copy of Exachk from <Document 1070954.1> into /opt/oracle.SupportTools directory on the primary node/LDOM of the SuperCluster.

    There may already exist an installed version of exachk, in which case the existing version should be backed up if desired or removed before installing the new version.
    There is a script that can be used to deploy Exachk in SuperCluster zones from the primary LDOM attached to this document. It will backup the current version, if any, and install the new.
    The latest version must be downloaded from <Document 1070954.1> and placed into the /opt/oracle.SupportTools directory along with the attached deploy_exachk.sh script.

2. Use the attached deploy_exachk.sh to deploy and install Exachk into /opt/oracle.SupportTools/exachk directory in the primary ldom and in each zone.

Example of what a zone's /opt/oracle.SupportTools looks like when this is finished:

3. Manual / Standalone installation

Manual installation can be accomplished as follows for each host / LDOM / zone in which exachk is to be installed

4. Check Version

5. Run it in GZ / Primary LDOM:

6. Run it in non-GZ local zone, aka NGZ:

Merging collections

With version 12.1.0.2.4 come some new features. For example, the ability to merge collections from various places into one file that can be uploaded to an SR via 'curl'.

Advanced Setup

The following is all about how to set up Exachk to run on a periodic basis.  First we set the following environment variables.

Set the autorun day of month and time of day based on customer input.

AUTORUN_SCHEDULE * * * * :- Automatic run at specific time

                 - - - -

                 ¦ ¦ ¦ ¦

                 ¦ ¦ ¦ +----- day of week (0 - 6) (0 to 6 are

                 ¦ ¦ ¦       Sunday to Saturday)

                 ¦ ¦ +---------- month (1 - 12)

                 ¦ +--------------- day of month (1 - 31)

                 +-------------------- hour (0 - 23)

Check the results

Run the presetup tasks. This sets up the required ssh equivalences.

Setup the daemon and inittab

Check that the daemon is running.

 Remember to keep Exachk up to date and to monitor root zpool file system usage on a regular basis!

References

This solution has no attachment