Network buffer used by Data Guard asynchronous shipping process

Question

Hi, all,
I am investigating the amount data that could be loss if we are using LGWR ASYNC for the redo transport services. I read somewhere it says the maximum exposure to data loss is the size of network buffer used by Primary database to ship redo apply. Where can I find out the size of this network buffer?
Thanks a lot in advance.

Answer 1

it says the maximum exposure to data loss is the size of network buffer used by Primary databaseIn 10.2 lns async reads directly from online log files, no buffer.
However, even with the buffer the above statement is incorrect.
There is no maximum to the exposure of data loss in either max performance or even max availability modes.
The exposure is a quantitative measure that depends on the speed of communication with standby.
You can evaluate it by comparing the rate of redo generation on primary ('redo blocks written' in v$sysstat) with that of redo apply on standby ('redo blocks read for recovery' in v$sysstat), or, easier, v$dataguard_stats on standby.
Oracle dataguard software sometimes [log switch] artificially increases immediate short-term exposure by parallelizing redo shipping.
But this parallelization makes it more efficient (and thus lessens this exposure) in the long run.
The only mode with maximum exposure limit is maximum protection, and that maximum is zero.