From OpenSCADAWiki
As a redundancy object of the subsystem "Archives-History" there used the object of messages archiver for which the redundancy process performs followed functions:
- Redundancy of the archiving mechanism — means of the messages combination from the redundant stations by the three mechanisms:
- in the main cycle of the redundancy there performs a request for new messages of the archiver on the master station from the last message time of the redundant object, where the last message is the last second of the time what is requested twice to prevent of lose messages on this time second, but there is a mechanism for controlling duplicates and suppressing them both immediately after the request and by the archiver himself;
- writing and rewriting messages to the redundant archiver causes to redirection of the writing request to to the master station both for new and rewriting messages (also the messages write locally);
- rewriting messages to the master archiver causes to the messages sending for all redundant stations, where is excluded of creation new entries into the current violations table — only modification of presented ones.
- Compensation of the data loss on the time of the node stop with the redundant node archive. There provides only initial synchronisation by loading/reloading of parts of the archive from the redundant station, at the station starting in general. The archive's part requested from the moment of the last record into the local archive minus value of the parameter of the depth of the forced reloading and till the current time.
Changed messages to more depth from the pointed will be lost on the started station! The initial synchronisation procedure involves in general:
- requesting for all active alarms;
- requesting for messages of concrete archive and on the depth specified by the parameter "Overtime of the reserve history reload at start" and up to the time of the previous request, so all new alarms won't definitely appear;
- coming to the normal mode of tracing new messages and alarms through the archive.
- Load distribution of the archiving between the nodes. When creating complex distributed systems there can be an important question of predicting and optimising of the overall system performance, taking this into the account the redundancy mechanism provides the execution of tasks of archiving of individual archivers only at one station. The other stations' tasks would go to the data synchronisation mode with the executive station. In the case of loss of the connection with the executive station the task of the local archiving is started. It's also provided the possibility of optimal the load distribution of the execution of archiving tasks of the archives group between the stations.
- Restoring of the alarms primacy. For the alarming messages, negative level, at their writing are performed an additional processing for the active alarms, that is only old-original ones are allowed for rewriting.