High-performance archiving of historical data enables complete traceability of system statuses. WinCC OA software offers two different archiving solutions: archiving in Value Archives (internal database format) or archiving in an Oracle database. The following documentation is provided as a best practice for monitoring the health of the WinCC OA connection/status to an Oracle database.
By default, WinCC OA gives you minimal information about the health of the Oracle connection or status. For this reason, adding alarm indicators to monitor the following scenarios;
- The database is no longer running or the Oracle computer is not reachable
- The database is no longer accepting data
Data logging failure
Oracle is a standard relational database used for long-term archiving and the centralization of data. WinCC OA in combination with Oracle provides customers with store and forward capabilities. This creates high reliability for your data. For example, if a network connection is lost, data can buffer and store locally on a WinCC OA server until the connection is re-established. Once reconnected, the data will forward to Oracle and fill in any missing data. This approach maintains a centralized data architecture.
The following is a project “config” entry example:
[ValueArchiveRDB] DbType = "ORACLE" DBUser = "PVSSRDB" DbPass = "Password1" Db = "WinCC" oracleClientVersion = 12 bufferToDisk = 2
The setting “bufferToDisk = 2” is the MAX buffer and will store every SQL command until the database is again connected. You should see this when it is working.
- The field “Buffer to disk path” can be set with config parameter “bufferToDisk”
- The field “Buffer to disk indicator” is essential. This indicates if your system is logging or buffering to disk
Monitor Status from datapoint element (DPE)
_RDBArchive.buffer.bufferToDiskIndicator (integer) normal mode = 0 building HD buffer = 1 reducing HD buffer = 2 Monitor this tag if there is an issue with the Oracle connection.
This is a discrete alarm type with 3 ranges.
- Value 0 is “normal mode”
- Value 1 is the alarm: “Lost connection to Oracle, buffering history”
- Value 2 is informational: “Oracle connection OK, backfilling history”
It is essential to know backfill (forward of data) is working because new data is not yet logging to Oracle. It must finish the backfill before it returns to “normal mode” and logging current value to Oracle.
An additional alarm that is important to monitor when the Oracle server is unreachable is a simple Boolean alarm on the connected tag.
In this case, a reverse bool alarm is in alarm when FALSE. Be sure to flip the “Good Range” and assign the alarm to the OFF condition.
Suggested Alarm Text
Connection to Oracle System Lost