We are in process of looking at using Oracle's Data Guard to replicate our 11g instance from a colo facility in Washington DC to Chicago.

To give some basics we have approximately 25TB of storage and a healthy transaction rate in the 1-2K/sec range. Also, because we are processing data in real-time we have a 24x7x365 requirement for processing data. We don't have any respites as far as volume except for system upgrades (once every few months) where we take the system offline but then course experience a spike in transactions when we bring the system back on-line. Ideally we would want the second instance in the DG configuration semi-online in a read-only fashion for reports/etc.

We evaluated DG in 10g and were not overly impressed and research seemed to show that earlier versions had issues with replication over a WAN but I have heard good things about modifications the product has gone through w/ 11g.

Can anyone confirm an instance of this size and transaction rate being replicated over a WAN and if so what is the general latency? An information or experiences with a DG implementation that is of this size and scope would really be helpful (or larger - I also realize we are still relatively small compared to many others out there).

Many thanks in advance.


I understand you're asking specifically about 11g but you may find the series of White Papers Oracle published on Maximum Availability Architecture (MAA) to be helpful. Specifically there is one called, "Data Guard Redo Transport & Network Best Practices" which you may find of use. I'm not sure if the series of papers has been fully updated but I suspect many of the points are relevant to 11g as well as 10g.



Data Guard Redo Transport & Network Best Practices MAA Best Practices - Oracle Database

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.