Hi
did you collect any outputs from the interfaces as the issues were occurring , any traces , interface outputs etc , could be many reasons but they would be the first things to check , without that information were just guessing really , traceroutes would be important for the latency so you could see the hop where the slowness occurred and then you could investigate that device or interface
As the device is new you could still check the interfaces you were connected to if they were only used for that once an see if there were any bad error counts on the ports but to check latency you need traceroute to see the path output