Natrinsic Supplemental Netezza Documentation

Shutdown or Decommission of Netezza Appliances

Preface & Disclaimer

This document provides a high level overview of the administrative and optionally physical steps to take when shutting down or decommissioning a Netezza appliance.

Natrinsic offers this guide only as a general overview; each customer may have unique requirements which cannot be covered in a general guide and as such Natrinsic does not warrant or guarantee the accuracy or completeness of this documentation.  Natrinsic will have no liability for any damages arising from the misuse of any information provided in this document.  Natrinsic also does not warrant or guarantee that any external links to third party websites will be active, correct or applicable for your situation.

Comments and Questions

If you require further clarification on the contents of this guide, you may reach us via email at: support@natrinsic.com with the subject line “General Support Inquiry”.  Within the body of your email, provide your company name, your position and the system identifier of at least one system for which you have a current support contract. (e.g. NZ123456)

Pre-shutdown tasks of the Netezza Appliance

Prior to shutting down your appliance, it is recommended to complete the following database related tasks.

Backups:

Server metadata 

nzhostbackup /home/host/nzhostbackup.<datatimestamp>.tar.gz  (system downtime is required)

Reference: https://www.ibm.com/support/knowledgecenter/en/SSULQD_7.2.1/com.ibm.nz.adm.doc/r_sysadm_nzhostbackup_cmd.html

Database backup 

nzbackup -history (ensure backup sets of critical DBs are secured on a SAN or NFS backup filesystem)

Reference: https://www.ibm.com/support/knowledgecenter/SSULQD_7.2.1/com.ibm.nz.adm.doc/r_sysadm_nzbackup_syntax.html

Audit

Audit database should be configured via '/nz/kit/bin/adm/nzhistcreatedb [options]' and DBA to review users' activities and historical query statistics

Reference: https://www.ibm.com/support/knowledgecenter/SSULQD_7.2.1/com.ibm.nz.adv.doc/c_advsec_audit_database.html

Shutdown of a Netezza Appliance - Platform tasks

Assuming that you have performed the pre-shutdown tasks, execute the following steps to shutdown and power-off 

1. Collect ‘head -25 /etc/hosts’ for IP addresses of HA1/HA2/VIP/HA1_IMM/HA2_IMM
   
2. ensure no jobs are currently running (review output of nz_responders)
   
3. nzstop (stop Netezza DB)
   
4. ssh 'ha2 service heartbeat stop' (stop standby host heartbeat service / assume HA2 is standby host by default if HA1 is standby host then stop heartbeat from HA1 first)
   
5. service heartbeat stop (stop primary host heartbeat service / assume HA1 is primary host by default if HA2 is primary host then stop heartbeat from HA2 after HA1 heartbeat service is stopped)
   
6. 'chkconfig drbd on;chkconfig heartbeat on' (turn off HA1 drbd and heartbeat daemon at run level settings)
   
7. ssh ha2 'chkconfig drbd on;chkconfig heartbeat on'  (turn off HA2 drbd and heartbeat daemon at run level settings)
   
8. ssh ha2 'shutdown -h now' (gracefully shut down standby host then check shutdown progress through IMM remote console if possible)
   
9. shutdown -h now' (gracefully shut down standby host then check shutdown progress through IMM remote console if possible)
   

Reference: https://www.ibm.com/support/pages/stopping-and-starting-ha-cluster-ibm-puredata-systems-analytics

https://www.ibm.com/support/knowledgecenter/SSULQD_7.2.1/com.ibm.nz.adm.doc/t_sysadm_power_off_n2001.html

Decommissioning of a Netezza Appliance

Assuming that you have completed the pre-shutdown tasks, and platform shutdown tasks, execute the following:

1. screen /opt/nz/fdt/storage_diags diskeraser --method clear (can take up to 12 hours per clearing pass)
   
2. screen /opt/nz/fdt/storage_diags diskeraser --method dod3pass (based on client's requirement)
   
3. nzinitsystem -reinit (reinitialize the system and restore the setting to manufacture default)
   
4. nzstart -newsystem (When the nzstart -newsystem brings the system Online the previous data on the SPU disks will not be seen)
   
5. ssh 'ha2 service heartbeat stop'
   
6. service heartbeat stop
   
7. pull out all the disks (containing /nz filesystem for catalog and several backup sets and log files) from both host server to scrap 
   

Reference: https://www.ibm.com/support/knowledgecenter/en/SSULQD_7.2.1/com.ibm.nz.adm.doc/c_sysadm_clear_data.html

https://www.ibm.com/support/pages/storage-data-erase-puredata-system-analtyics

https://www.ibm.com/support/pages/how-do-data-sanitization-ibm-puredata-system-twinfin