- Published: September 26, 2022
- Updated: September 26, 2022
- University / College: The University of Newcastle, Australia (UON)
- Language: English
- Downloads: 8
On December 16, 2006 the migration of the production database will take place. It has been deemed necessary to move our Oracle production database to a more robust server for future growth. At this time the database resides on a p650 unit that will be utilized in phase two of this migration. Below find the laid out plan as it pertains to the move. December 16, 2006: There will be a complete backup of the existing server. December 17, 2006: Starting early that morning, the backup of the existing server will be restored on to the new server emulating the currentenvironment.
While the restore of the new server is taking place, the existing server will be set up for the actual move of the existing Oracle database. Step. 1: The mounted file system will be un-mounted from the operating system, at that point after recording all the necessary paths. Step. 2: The volume group that the file systems reside in will be varied off, and then exported. After this step the database which resides on the EMC Symmetric will no longer exist on the present database server. Step. 3: The server will be totally taken down and power will be totally extinguished. Step. : The remove of the fibber channel card that leads back to the EMC database location. The data on the EMC/database will go unchanged or accessed at this point. Step. 5: The fibber channel card will then be placed in the new server. Step. 6: The new server will then be powered on with the fibber channel card from the now existing server, and then the database will be imported to that server. Step. 7: At this point all the network and fibber connections will be moved to the new sever from the existing one using the same IP/Hostname to eliminate further configuration changes or delays. Step. : A DBA will be contacted to confirm the migration success, and then the server will be taken down for the data center outage. System Upgrade Recommendation Phase II Oracle Environment As a result of an evaluation of performance and future growth of the present Oracle environments, it was proposed and recommended that the production database server be upgraded with additional processors and memory. As phase one, on January 29th 2006 each Oracle environment received the recommended upgrade as described in the briefing dated January 3, 2006 and title: Hardware Upgrade for Oracle Financials in Preparation for OAB and iSupplier Modules.
The Oracle environment received the recommended memory and has proved to perform as projected. The production application server now utilizes 16GB of memory, and the database server has a total of 32GB. In a continued effort to move toward phase two, the following information has been composed and now submitted for evaluation. It is the opinion of the system administrator, utilizing performance tools that the existing system degradation only appears during multiple thread requests.
In short, thread request are granted buy and throw system processors within any given computer environment. To wit, it has been ascertained that during a normal business run the system reflects system degradation at any point of a new presented workload introduction. System memory and IO performance proves to be stable while the now utilized processors show to be at their maximum thread capacity. In today’s environment the production database unit consists of four processors; with the recommendation to increase this number buy four additional processors facilitating a total of eight.
This is projected to improve Dallas County production presently as well as future growth performance. See projected cost on page two. Model Highlights 7038-6M2 The Model 6M2 delivers a cost-efficient growth path to the future with: • 64-bit system scalability in 2-, 4-, 6-, and 8-way configurations with the following processor options: o 2-way 1. 2 GHz POWER4+ with 8 MB shared L3 cache per processor card o 2-way 1. 45 GHz POWER4+ with 32 MB shared L3 cache per processor card • Expandability and reliability: System memory expandable from 2 GB to 64 GB o Rack-mounted drawer utilizes 8U (EIA Units) o Supports up to eight 7311-D10 or 7311-D20 I/O drawers per server o Each I/O drawer supports either 6 (for D10) or 7 (for D20) hot-plug PCI bus slots To upgrade to an 8 way system IBM has given a cost of $25, 750. 00 2-way 1. 45 GHz POWER4+ with 32 MB shared L3 cache per processor card. The above quote is without any government discount or third party intervention. However, a third party vender’s quotes $9590. Mr. UNIX Sr. UNIX System Administrator