Spotlight on Oracle 11.0 - Getting Started Guide

This is where the flow of the Spotlight on Oracle Data Guard home page starts. The Redo Log Writer writes Redo Buffer Blocks to the Redo Logs.

The Redo Log Writer also writes to the Log Network Server. If the Transport Method is Synchronous (or Parallelsync) then a bottleneck at the Network or Standby could prevent the Log Network Server from flushing. The Redo Log Writer would write the overflow through to the Redo Logs to be transferred to the Standby via the Archiver.

Note: A number of alarms are associated with LGWR on Spotlight on Oracle. Those alarms are raised when applicable on the Spotlight on Oracle home page. They are not raised on the Spotlight on Oracle Data Guard home page.

The overhead of DataGuard on the Primary database.

The design is similar to the Average Active Sessions (AAS) gauge.

• Dark Shaded Area - Amount of time sessions spend waiting on data guard.
• Light Shaded Area - Amount of time sessions spend on other tasks.
• White Line - CPU Limit.

Either of the colored bars passing the vertical white line indicates potential performance issues. How far the colored bar passes the vertical white line indicates the severity of the performance issue. A value far greater than the CPU limit indicates that the database is experiencing bottlenecks.

The Data Guard Overhead Ratio Alarm is raised when a high proportion of the Primary's activity time is spent in Data Guard activities.

Is the Data Guard Broker enabled?

The Data Guard Broker is a utility designed to monitor and manage Oracle Data Guard. It can be setup to automatically failover to the Standby if the primary becomes unavailable.

Transport Lag impacts your business Recovery Point Objective. You want minimal transport lag so the Standby has all the information it needs to become the Primary (with minimal data loss) should the Primary go down.

Set the Transport Lag Alarm thresholds accordingly.

Notes:

• There will always be a Transport Lag when the Archiver is used. The Archiver waits for the Archive Log to fill on the Primary before it transports the log to the Standby. Via the Log Network Server changes on the Primary are transported to the Standby as they happen. If there is a bottleneck on the Log Network Server then the Redo Log Writer will write the overflow to the Redo Logs to be transferred to the Standby via the Archiver. A bottleneck on the Log Network Server could be caused by a slow Network connection, particularly if the Transport Method is Synchronous or Parallelsync.

• In Oracle versions prior to 11.2 some configurations experience inconsistent or misleading data on the Transport Lag. This is documented as Oracle Knowledge Article [ID 1299000.1].

  An Example Scenario: Outstanding Logs are speedily applied. Therefore you would expect the Transport Lag to be low, but the Transport Lag is increasing. The Transport Lag data is misleading. Disable the Transport Lag Alarm and use the Outstanding Logs to Ship Alarm to check for bottlenecks.

The method of transport used from the primary to the standby.

• Synchronous or Parallelsync: The Log Network Server ensures its data is received on the Remote File Server before it clears its buffer. This can lead to extra load on the Primary. This can be gauged on the Data Guard Overhead as per the Data Guard Overhead Ratio Alarm.
• Asynchronous: The Log Network Server clears its buffer without checking whether its data is received by the Remote File Server. This puts the least load on the Primary but impacts business Recovery Point Objective. When the Transport Method is Asynchronous the Transport Lag should be zero.

The number of Log Network Server processes on the primary.

The Transport Process Running Alarm occurs when there are no Log Network Server processes on the primary.

The number of Archiver processes on the primary.

The rate at which LGWR processes write entries to the Redo Logs.

The rate at which log files are copied to the Remote File Server through the Archiver averaged over 5 minutes.