![]() | Note |
---|---|
With V10, database queries can also run via the PARTapplicationServer. In such a scenario, replicating the LinkDB is not necessary. See Section 1.3.5, “PARTapplicationServer installation/configuration”. You can find a diagram for multisite installations at Section 1.3.5.2, “Overview Multi Site Installation ”. |
For distributed locations, it may make sense to replicate the LinkDB. Reasons may be that the transfer times are too long; or the need to be able to work autonomously and the system availability.
Data replication is the copying of information to one or more databases in a manner in which the data is equal in all locations. There are basically two different types of synchronisations: synchronous and asynchronous replications. With synchronous replication, the data comparison occurs nearly without any delay. With asynchronous replication, seconds, minutes and possibly even days may go by. In addition, it is important to know whether all so-called subscribers only have reading rights, or whether they are also able to change data.
The current replication tools from Microsoft and Oracle, as well as other providers, are adjustable to special requests and offer plug & play solutions, which can be used with relative little implementation complexity.
You can find detailed information on the subject of replication on the Microsoft and Oracle websites.
![]() | Note |
---|---|
Please also note the Section 4.2.1, “Hardware requirements ”. Data throughput per time unit and the part type have a decisive influence on performance. |