I have a CCR1036 that has a 100 megabit/s circuit behind it. When doing an iperf and filling the pipe (85Mbit/s queue), Winbox is transferring 6.3Mbit/s across that same WAN pipe. I find in troubleshooting, I have to carve out a larger than expected slot for Winbox to not crap out as my customers are using the pipe to it's fullest.
Can anyone shed light on the reason why Winbox gets chattier when the pipe is full? When I open up multiple windows to graph, I could see it reaching to the device for data more often - but I tested with all windows closed except for the firewall rule page (which is - sad to say, not very populated; but it is a point to point.)
I only mention this because the service provider has a much harsher cap on that 100meg than I do - and it does throw Winbox for a fit. I just didn't expect to see a relationship between circuit utilization and 8% increase just because I have Winbox attached to it.
Can anyone shed light on the reason why Winbox gets chattier when the pipe is full? When I open up multiple windows to graph, I could see it reaching to the device for data more often - but I tested with all windows closed except for the firewall rule page (which is - sad to say, not very populated; but it is a point to point.)
I only mention this because the service provider has a much harsher cap on that 100meg than I do - and it does throw Winbox for a fit. I just didn't expect to see a relationship between circuit utilization and 8% increase just because I have Winbox attached to it.
Statistics: Posted by mlacomb — Sat Feb 17, 2024 4:54 am