Having a confusing problem using a backup to restore a CA on another machine.
I’ve created a backup using the manual option in SMART desktop and shared this file with a colleague. He used the Advanced option at login and selected option Update/Restore from backup and this all proceeded fine. But when he tries to sync to Connect he gets an error as if there is something broken in the database:
org.apache.derby.shared.common.error.DerbySQLIntegrityConstrinatVioloationException: DELETE on table ‘PAROL_FOLDER’ caused a violation of foreign key constraint ‘PAROL_FOLDER_UUID_FK’ for key ( long key number). The statement has been rolled back.
This is the second back up file that I’ve sent him that produced the same sort of error. Could it be that the backup process is “breaking” the database in some way? The source machine (mine) syncs with Connect fine and database seems to be perfectly fine and functional.
I haven’t tried doing a restore from the back up on my own machine. I’d be slightly nervous to try.
My reason for asking him to use a backup rather than download the CA from Connect is because the file to be downloaded from Connect is 4GB (for some unknown reason) whereas the backup file is only 700mb. Perhaps this signals an issue somewhere. But I don’t have control over the Connect database.
When using SMART Connect, it is important to manage the distribution of data through Connect rather than exporting/importing between desktops. This maintains the unique identifier (UUID) that SMART uses to track CAs, and importing the CA separately can cause it to use a different UUID which break that linkage with Connect.
That’s important to know Matt! Thank you. Now all the more urgent to resolve the problem we have with a bloated CA on Connect making it so hard to use that route. I think the bloating is due to unprocessed patrol packages that seem to get bundled into the download from Connect. Maybe… Jeremy
When I download our CA from Connect it has c 16,000 old data packages in the smart_connect filestore. I’ve carefully gone through them all (yes!) and retrieved lots of stuck data and have now deleted all the stuck message we had in the data queue (about 800 of them - so far fewer than are appearing in the filestore). But if I download the CA again the filestore still has all these old unneeded messages in it.
Is there a way to clear this out on Connect? They cause the download to be over 4GB whereas the full backup is just over 1GB. My instance is hosted by the partnership so I don’t have developer access to the database.
Many thanks for any suggestion.
The solution is really to delete all the old entries from the data queue. If you want I can do that for you with a small tool I created. I only need a login on your SMART Connect server and some guidance, which entries can be removed from the queue.
Hi Oliver thanks I’ll email you the login details offline. I’m not sure what queue you mean. The data queue that I can see on Connect is empty. I deleted everything hoping it would solve the problem. Patrols are paused at present so no new packages are coming in.