9.3. Troubleshooting
A good first step in troubleshooting redundant DataHub connections is to check the Event Log. There you may see error messages related to a configured redundancy pair. In addition to this, the Redundancy feature offers some useful troubleshooting options.
Diagnostic Points
In the Redundancy Configuration dialog you can specify diagnostic points that will get added to the root of the output data domain. These points let you monitor the status of the redundancy pair.
![[Note]](images/note_1156523861.png) | |
These points will continue to update even when redundancy is disabled. This gives you a way to examine the control bits to determine that the redundancy pair is not currently enabled. |
There are two kinds of diagnostic points.
Status and Control Data Points: Assigning names to these points will add them to the root of the output domain. If you start each name with a unique character or two, like a double underscore (
__), the points will all sort together at the beginning or end of the point list, making them easier to find and view together.
Point for current source number: When set, this point indicates which source domain is currently in use. Its value will be one of these:
0 for none (neither input domain passes the validity check).
1 for source domain 1.
2 for source domain 2.
Point for current state of domain 1: This point's value will be one of:
0 if the data in domain 1 is not valid, according to the validity settings.
1 if the data in domain 1 is valid.
Point for current state of domain 2: Same as above.
Point for preferred source number: This point's value will be one of:
0 if there is no preferred source.
1 if input domain 1 is the preferred source domain.
2 if input domain 2 is the preferred source domain.
Input Statistics Data Points: In the Options section of the Redundancy Configuration is an option to Create data points showing input statistics, which is off by default.
If this is turned on it will create the following data points in the output data domain:
Input1ValidCount shows the number of valid points in input domain 1.
Input1InvalidCount shows the number of invalid points in input domain 1.
Input1UninitializedCount shows the number of uninitialized points in input domain 1. Uninitialized points are ones that have not yet received a value from any source. Changing the radio-button selection from Invalid to Valid tells the redundancy engine to ignore such points when evaluating the validity of the domain. Selecting Normal treats them like all of the other points in the domain.
Similar points are created for Input2.
In addition to these, for OPC A&E or A&C connections a connection status point is available for monitoring the connection. Please see A&E / A&C Connection Status Point for more information.
Forcing a Switchover
The DataHub Redundancy feature will always attempt to maintain a valid output domain by selecting one of the valid input domains. This means that you cannot force a switchover to an invalid input domain. You can request a switchover by changing which input domain is preferred, and the DataHub instance will comply if that preferred domain is valid.
You can change the preferred input domain by changing the value of the
Point for preferred source number control point (see above)
between 1 and 2. For example, if you have
named that point _MyPreferredSource, simply change its value
through the Data Browser, an attached DataHub client like a GUI, or a script. If the
preferred source is domain 1, enter the value of 2 to change the
preferred source to domain 2. The DataHub instance will immediately switch to domain 2,
as long as that domain is valid.
| |
Changing the preferred source in this way does not alter your DataHub
configuration file. Also, using a value other than |
Switchover Pause
Any pause during switchover is the time it takes to confirm that the current
source domain has become invalid. Normally this is the connection timeout period.
Until that timeout occurs, the DataHub instance still considers that domain to be
valid. The status for those points won't change to Not Connected
until the DataHub instance is aware that the connection has been lost. To reduce the
time of the switchover pause you can reduce the connection timeout on the sources,
or configure Data Flow Detection for this redundancy
pair.