I have a synchronization problem with synchronizing two ASA databases. I have to synchronize some tables (download only), and one of the table will have big amount of data to synchronize (about 150.000 – 300.000 rows ). The synchronization process takes forever.
As far as I understood, the mobilink server works as follows:
1. gathers all the data that will be sent to the remote database
2. sends all these data to the remote server at once
3. the remote database struggles to add these data
When the first step occurs, the processor goes up to 100% and stays there till the data gathering finishes. This happens when only one remote database synchronizes, what will happen when several remote databases will try to synchronize with the consolidated database?
Now, the questions:
It is the above description correct?
There are some wais to optimize this synchronization?
1. ml_user client prepares/sends an upload stream to sync server
2. Server prepares/sends a download stream to the remote
3. client confirms reciept.
Now coming to your problem with CPU's. Mobilink is light weight and completely operates using via DSN's. The streams are not cpu bound rather io bound. So, I suspect this may be some bug in the threading or connection pooling. I say maybe, I haven't seen this happen. I would check with the iAnywhere tech support.