2012-04-23 01:00 AM
I am working to setup OCI in my environment. According to the infrastructure, I prepared sizing of 2 servers, 1 DWH and 2 RAU, located in different DC but they are interconnected to each other.
My question is how to configure servers to work independently or to manage separate storage systems or switches?
Like if I configure 1 core switch, OCI automatically discover all connected switches or configuring a command suite detects all HDS systems.
How can I create a logical separation between them or configure in a way they won't discover or collect duplicate data?
2012-04-24 10:44 AM
There are several ways to address this. With all the data sources you can use exclusion lists. With either Brocade or Cisco you can turn off the discovery option and create only one data source to one switch.
The architecture questions are best addressed through the sales channel. Support does not offer architecture or design suggestions, just tells what the capabilities of the system are. Important point it to only gather data from each source once. Redundant data collection puts a higher stress on the devices in your production environment and may create issues with collecting and data normalization.
Please contact your sales team.
2012-04-24 11:09 PM
I have some query about tiering rules.
I have storage systems like HDS (VSP, USP, AMS etc) and IBM (DSXXXX, XIV, SVC) and behind some of them there are different types of disks FC, SATA, SAS.
1. Now when it comes to define tier, how do I differenciate them? (A single system have combination of FC, SAS and SATA)
2. Is there any way of auto tiering?
3. What are standards of different tiers like T1, T2, T3? Is it based on RPM too?
2012-04-25 06:45 AM
Hello, Sunil Yadav.
You can find information about how to do auto tiering in OnCommand Insight 6.3 by going to the NetApp Support site and looking at the OnCommand Insight 6.3 Inventory User Guide.
Here's an excerpt from that document.
Configuring volume service level rules for automatic assignment
To improve storage efficiency (defined as cost per TB at a certain IOPS level), a volume (or LUN) can be assigned to storage levels dynamically using automatic storage tiering policies. Based on policies you configure in your device, automatic storage tiering continuously migrates blocks from one service level to another when those blocks are more or less frequently used.
Automatic storage tiering is a storage technology that uses more than one physical tier of storage as a single LUN. This technology is also referred to as dynamic tiering, full automated storage tiering, sub-LUN tiering, or hybrid aggregate tiering.
In this document, you'll find information about Flash Pool tiering and FAST VP tiering.