This content has been marked as final. Show 3 replies
Fabricio_Jorge wrote:For maximum protection it is recommended to use at least more than one standby database, So that synchronously it can write at least on one standby. so you must have either good band width or distance should be less, In that case if you have two standby one can be near where there is no network problems and for high availability it can be other location.
I would like to hear (or read) from you all, how actually find out if the network is a "problem" in a data guard configuration. For example, if a have two sites, SITE A (primary) and SITE B (standby away from site a with a distance of 20 miles, for example). This configuration is using SYNC redo log transmition (Max protection).
The main doubt here is: With the configuration above explained, how to find if i have a network problem, regarding slowness? For example, does AWR report something related to this? What OS tool should i use to discover this slowness (ping?).
Thanks in advance.
In AWR reports you can see the wait class type related to network also, But AWR can give you most information from the database. Not from the OS level network latency, port latency, ping packet lost so on.
You can also work out on SDU/TDU as mentioned by Mseberg, You can check some good information on Redo transport and network best practices from http://www.oracle.com/technetwork/database/features/availability/maa-wp-10gr2-dataguardnetworkbestpr-134557.pdf