How to sync a huge table in batches to prevent OOM errors

[warp10]

I need to sync a huge database table from Odoo (it contains a binary field with a lot of data), and unfortunately we need to sync it all at once, as we must have the data available on the device for offline use.

Now, our problem is that we get an OOM error when we try to sync the table from Odoo to Android, and I suppose it is related to the size of the table itself. I read an interesting hint by @dpr-odoo on Issue #216, that is using syncDataLimit() to limit the amount of data synced, so to prevent the OOM error. My question is: how can I reiterate the sync in batches until the whole table is properly synced?

[adam-nadar]

try using ir.attachment to download huge binary data referencing it..... so get the binary data when it is needed and update the record with background service..

[kasim1011]

@warp10 Hello, you can always register receiver for ISyncFinishReceiver.SYNC_FINISH on OnResume(). after you do this when you start sync for any model the iSyncFinishReceiver->onReceive() will call when syncing get finish for that model. so, you can start syncing for next model.

private ISyncFinishReceiver iSyncFinishReceiver;

@Override
protected void onCreate(Bundle savedInstanceState) {
    super.onCreate(savedInstanceState);
    iSyncFinishReceiver = new ISyncFinishReceiver() {
        @Override
        public void onReceive(Context context, Intent intent) {
              // previous model sync done. request for next model sync.....
        }
}

@Override
protected void onResume() {
    super.onResume();
    this.registerReceiver(iSyncFinishReceiver,
            new IntentFilter(ISyncFinishReceiver.SYNC_FINISH));
}

@Override
protected void onPause() {
    super.onPause();
    try {
        this.unregisterReceiver(iSyncFinishReceiver);
    } catch (Exception e) {
        e.printStackTrace();
    }
}

[mikedream89]

@warp10 Do you fix about sync a huge database table?

[warp10]

@mikedream89 Not yet. The suggestion from kasim1011 doesn't really apply to our case, as we don't have any issue in starting another model after the current one, we have problems syncing one big, specific model. We worked around that limiting the number of items synced at each sync, but that's not a actual solution