Attached is what our network performance looks like over an 8 hour aux copy period. Data was moved for about an hour while the job ran for 8 hours and was killed by the schedule (19:1111) unfinished. Seem like Aux Copy is doing something other than moving data. The network has no bottlenecks or policies. From time to time when I see this as it's happing, I can manually copy data to the source at expected rates.
We haven't scheduled the on site visit yet. I'm hoping that will be done sooner than later.
To date support has suggested these 4 items be changed.
- Distribute data evenly among multiple streams for offline read operation
- bypass wan optimization devices.
- Increasing the Speed of the Read Operation
Enable Dash Copy/Network Optimized copy
We applied the final pieces today. We should know the results shortly.