0
votes

I've been working with the Sync Framework from microsoft with c# trying to synchronize about 35 tables from a local database to a database stored in a central server, the main problem is that one of my tables has more or less than 1million records, and even with the filters it takes to long to synchronize, i dont know if there's a way or any other framework that works a bit faster than this.

for the complete sync it takes about 4-6 hours.

any help will be good, thanks in advance.

3
welcome to SO! it's generally recommended to show some code with what you've already tried. possibly, you just need to use the existing framework a different way. - Corley Brigman

3 Answers

0
votes

Sync framework is just like any other database app and you can troubleshoot/optimize it similarly.

You can enable sync framework tracing to see where its spending its time: querying for changes, serialising changes, applying changes, locks/concurrency issues, network latency, etc...

Do you have pre-existing data on the destination DB? Have you tried initializing replicas from backup? Have you enabled batching? etc...

0
votes

I am not sure if you still looking for a solution, but while using sync select upload / download incremental changes. This way every time you sync only new changes will be transferred. This way the process will consume time on the first sync only, later on only changes will be uploaded / downloaded, this it will be fairly quick.

0
votes

If it has million records it will take lot of time to sync...Better is to build ur own sync architecture...we faced same issue then we created our own framework with rest based web apis returning Json formats.Only last updated rows will be synced based on last updated time...dnt need to compare whole database...