Thanks. To follow up in case anyone else comes across this, it seems to be an issue with our hosting environment. It’s running in a container with a virtual file system and I think it’s just not able to keep up with the read/write requests so things get locked. Engineers are investigating.