Restore Speedtest Frequency Options
M
Mark Murphy
Wattbox's will ping a WAN or LAN IP address or URL's every 15 seconds and display a percentage timeout in OVRC's device page. Most people setup a reboot of the internet modem, router, network switch, and other monitored devices based on this information. Its by far the most useful tool for determining internet and local network communication problems via OVRC. Granted they are called ping 1, ping 2, ect by default, but renaming them and adding what path you want to monitor via a ping is easy enough. The total time period tracked for % is not changeable and I think its only the last hour or two which is short for site history review. Also % timeout is not representative of high latency periods, but hopefully OVRCs new latency monitoring improves upon this concept. The new latency test for WAN/LAN once every hour is redundant to what wattboxes already do, but not frequent enough to be useful. Speedtest should be thought of like an internet STRESS test and not something you would want running frequently to diagnose and determine internet problems. It uses a lot of data and It would become a potential hinderance to the normal internet traffic.
OVRC depends on the sites internet to be working to display up to date information, so at best the information about the status of the sites network will have a small delay, at worst you have to manually do a network scan and wait a few minutes for updated information. This delay could be monitored and displayed to indicated the sites internet quality as well.
Wattboxes can monitor voltage and current usage as well. The history of that being displayed in OVRC would be a nice tool.
C
Chris Fitton
Mark Murphy: Not sure I correctly understand your response regarding speedtest, but it sounds like the logical conclusion of your statement is that speedtest feature is not worthwhile.
The problem speedtest identifies, as I use it, is when the ISP-promised delivered speeds OVER TIME (days and weeks) is not met. This is "felt" by customers' or my perceived slowness. Unfortunately, in my area, the service (speed) is all over the place. This evidence can be used to "help" the ISP better troubleshoot beyond rebooting the modem or router.
Yes, a speedtest running every minute continuously COULD be a hindrance to users, but a single test every hour by itself won't make a difference. OvrC went from a max 4 tests per day to 1, rendering its usefulness to nearly null.
M
Mark Murphy
Chris Fitton: A standard ping cmd sends a packet about 256 bits in size and gets a packet 256 bits in return. It also returns a round trip time. This gives you your base speed as a minimum data/ minimum time rate. It can be monitored every second, or quicker, with no impact on other traffic as it is measuring a minimum. If you monitor this you can ensure communications between devices doesn't go above a desired base. Knowing its a minimum amount of data, many software monitoring ping commands by collecting massive amounts of them and show statistics min, max, and average of the time not a data rate. Ask any kid who plays online games, latency(ping) is the most critical parameter to their online experience. This is true for all network traffic. A speedtest is just the opposite a much larger data size which is downloaded and uploaded, taking up a much longer time period to give you a megabits per second rating. Basically clogging up the entire download and upload bandwidth in order to tell you the currently available bandwidth. If you have done speedtest on a busy network with many other devices using the bandwidth you may have notice you don't see the full bandwidth, just what is available to your device. You need to have only one device connected to the ISPs modem doing nothing but running a speedtest to get a true available bandwidth rating. All other speedtest are prone to not seeing bandwidth because its being used by other devices. Now if you are following along, what you can do is measure the effect bandwidth usage has on you minimum base latency. This is a true test of how well a network connection appropriately handles traffic or falls short of the claimed bandwidth speeds at high usage. The best part is this doesn't have to be data downloaded and uploaded by a speedtest website, but by the natural traffic from users of the network.
Service AVX
1000% agree, it's practically useless now. If my client doesn't have internet, they will tell me a lot faster.
K
Kit Dunagan
Even 4 per day was not enough resolution to determine if intermittent speed and connectivity issues were time of day dependent. Bakpak from pakedge allowed you to set a speed test frequency to multiple times per hour. Even as a short term troubleshooting setting it was very helpful.