RE: Exadata Machine spare part administration

When you purchase an Exadata you get a base spare parts pool per rack (1 CELL disk drive and 1 Flash Card).

Yes if you believe that you need a larger spare parts to include other components then you have to purchase the spare parts.  

As to redundant, at the fundamental part level all parts are not redundant within a specific server/switch, but at the Structural component level Compute Node/CELL Nodes you are redundant. At the Compute node that redundancy is only in play if you are using RAC instances on the Exadata.  

I would never agree with this statement: “make sure we will not have any downtime in the production environment”

I have seen

1. Some setups that have single instances without failover setup under GRID so if you lose a compute node your instance is down.
2. Regular data center work on the Power distribution system where half the power is taken off line which for one part of the data center caused a power surge to the secondary power distribution and took out multiple racks.
3. No matter how many spare parts you have it will not keep you from having a multipart failure like losing 2 disk drives at the same time under Normal redundancy causing ASM disk group loss.
4. The standard bugs we all hit that actually cause the clusterware to bounce multiple or all compute nodes.

It really comes down to simply the statistics of risk calculation, but there is no guarantees of “never having downtime” no matter how many levels of redundancy you have.          

From: oracle-l-bounce_at_freelists.org <oracle-l-bounce_at_freelists.org> On Behalf Of Kamran Agayev (IT/DBA) Sent: Thursday, November 28, 2019 6:42 AM To: oracle-l_at_freelists.org
Subject: Exadata Machine spare part administration  

Hello everybody  

This year we were informed that Oracle doesn’t allow partners to have a stock for Exadata spare parts. It means that you have either order a spare part once it’s failed (which takes time for the shipping etc.) , or buy and keep it in your stock. They just don’t provide extra spare parts (for free, to keep it in partner’s stock)  

As almost everything in Exadata is redundant, it means we can order the failed hardware and just wait for it for few days and then use it once it’s arrived. We have hosted some critical production databases in Exadata machine (X7-2) and we also have a fully functional standby Exadata. However, to maintain 24x7 availability of databases, we have an option to buy some spare parts in advance and keep them in the stock and make sure we will not have any downtime in the production environment  

So I have 2 questions:  

• Can we confirm that the Exadata machine is redundant by itself and we don’t need to buy spare parts in advance? (with and without considering the standby Exadata, as it’s going to take some time to perform switchover of all databases)
• Did you have any bad experience regarding Exadata hardware failure which caused downtime of whole Exadata machine so far?

Thanks beforehand

BR,  

Kamran Aghayev A.  

This e-mail and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you are not the intended recipient you are hereby notified that any dissemination, forwarding, copying or use of any of the information is strictly prohibited and the e-mail should be deleted immediately.

Bu e-mail və onunla göndərilən istənilən fayllar məxfidir və yalnız ünvanlandığı fərdin və ya müəssisənin istifadəsi üçün nəzərdə tutulmuşdur. Nəzərinizə çatdırırıq ki, əgər Siz nəzərdə tutulan qəbuledici deyilsinizsə, göndərilən məlumatın hər hansı şəkildə yayılması, yönləndirilməsi, kopyalanması və ya istifadəsi qəti qadağandır və həmin e-mail dərhal silinməlidir.

--
http://www.freelists.org/webpage/oracle-l