Amplify Sync Expression - to sort by provided field

[manbalagan]

Problem : For chat application - for example, has 2 contacts with 5,000 messages each. If the sync expression syncMaxRecords is set to 5000, then messages from first contact gets synced and not for the second contact.

Expectation : Messages should be synced based on creation date (SortKey), so that latest 5000 messages gets synced irrespective of contacts.

Sync Expression:

DataStoreSyncExpression consumerDataSyncExpression = () -> ConsumerData.INCARCERATE_ID.eq(siteIncaratedId); DataStoreSyncExpression textMessageDataSyncExpression = () -> TextMessage.INCARCERATE_ID.eq(siteIncaratedId); DataStoreConfiguration datastoreBuilder = DataStoreConfiguration.builder() .errorHandler(STCDataStoreErrorHandler.instance()) .syncExpression(ConsumerData.class, consumerDataSyncExpression) .syncExpression(TextMessage.class,textMessageDataSyncExpression) .syncPageSize(1000) .syncMaxRecords(25000) .syncInterval(15, TimeUnit.SECONDS).build(); AWSDataStorePlugin dataStorePlugin = new AWSDataStorePlugin(datastoreBuilder); Amplify.addPlugin(dataStorePlugin);

GraphQL Schema
schema.zip

Exception
DataStoreException{message=Initial sync during DataStore initialization failed., cause=java.util.concurrent.TimeoutException: Timed out while performing initial model sync., recoverySuggestion=There is a possibility that there is a bug if this error persists. Please take a look at
https://github.com/aws-amplify/amplify-android/issues to see if there are any existing issues that

Replying to Closed issue
Can I get code sample of what your graphQL schema and the sync expression looks like? I am under the impression that it is supposed to sync based on a filter irrespective of which contact a message belongs in.

Originally posted by @raphkim in #1430 (comment)

[poojamat]

The data coming back from the sync query is ordered only by the hash key. Unfortunately we do not support specific sort during the sync operation. I will add this as feature request.

[manbalagan]

To overcome the time out - Is there a specific API that can override the time out of ~1 minute.

Currently I am doing a work around based on network status on datastore when the connection times out

    Amplify.Hub.subscribe(
            HubChannel.DATASTORE,
            hubEvent -> DataStoreChannelEventName.NETWORK_STATUS.toString().equals(hubEvent.getName()),
            hubEvent -> {
                NetworkStatusEvent event = (NetworkStatusEvent) hubEvent.getData();
                if(event.getActive()) {
                    Log.i("xxxx", "Datastore sync isActive ? : " + event.getActive());
                }
                else{
                    Log.i("xxxx", "Datastore sync isActive ? : " + event.getActive());
                    xxxx.runOnUiThread(new Runnable() {
                        @Override
                        public void run() {
                            isRetry.set(true);
                            loadContent(); // calls the Amplify.Datastore.query
                            Log.i("xxxx", "Timed out - Retrying....");
                        }
                    });

                }
            }
    );

[poojamat]

Unfortunately there is no way to set this time out at this time. We will put this is a feature request as well.

[manbalagan]

Is there an updated datastore library released for this feature request for syncing based on given criteria.

Expectation : Messages should be synced based on creation date (SortKey), so that latest 5000 messages gets synced irrespective of contacts.

[KevinLam0308]

I'm facing the similar issue and used the same workaround here, the sync time is over a minute at the first time syncing, here is the reported issue:

https://github.com/aws-amplify/amplify-js/discussions/9680

Ultimately, set a correct syncExpression to minimize the sync time and reset the timeout during syncing should be the best approach, please help to put this as a feature request also, thanks!

[manbalagan]

@poojamat Is there an issue created for taking this as a feature request ? Do you have a timeline on which version this would be released ?