Database Reference
In-Depth Information
Figure 14-15.
AWR ping statistics
This helps determine the latency encountered at the database level vs. any overheads at the operating system level.
Apart from the basic packet transfer tests that can be performed at the O/S level, there are other checks and tests
that can be done to ensure that the cluster interconnect has been configured correctly:
1.
There should be redundant, private, high-speed interconnects between the nodes
participating in the cluster. Either implementing NIC bonding or pairing or by configuring
multiple interconnects (HAIP) in Oracle Database Version 11.2.0.2 will help provide load
balancing and failover when one of the interconnects fail.
2.
The user network connection does not interfere with the cluster interconnect traffic,
keeping the network traffic isolated from each other.
Conclusion
In this chapter, we looked at why performance problems centered around the interconnect may not always be related
to the reasons why RAC cannot be the solution to highly available machine critical scalable architecture. Most of the
interconnect-related issues are centered on bad configuration and/or bad and faulty interconnect infrastructures
such as switches, NICs, and so forth.
With the technology advancement and improvement in networking configuration options available from the
industry, high-speed interconnect options are possible for applications such as data warehouses where there is a high
volume of data that is transmitted and where throughput is a concern, and high-speed interconnects such as 10GigE
and Infiniband options are available.
Oracle also has been making considerable improvements on the configuration options, providing scalable
interconnect options with the HAIP feature in Oracle database 11g Release 2 (11.2.0.2).
Search WWH ::
Custom Search