What's new

spdMerlin spdMerlin - Automated speedtests with graphs

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

With more processes being run on the router particularly SPDMerlin is there a way to include the router as a client when looking as the statistics under this section?

The reason I ask is based on an analysis I did and the gap between data use in traffic analyzer by day and then the data use as accounted for by device has increased dramatically

My comparison is based on three days data use both up and doen before installing spdmerlin and connmon and after.

Before: Total Use = 49.45 GB By Device 46.54 GB = 94% accounted for by device

After: Total Use = 54.02 GB By Device 19.28 GB = 36% accounted for by device

If someone else has an explanation for this increasing variance please post it.

I am running amtm, Entware,Skynet, SPDMerlin and commcon on an AC86 running Merlin 10.2
 
the obvious explanation i can find is the connmon ping and the spdmerlin test use data. (probably spdmerlin more so) if you do a normal speed test on like lets say ookla for example you are using data to do such. i have my connection dropped down to 250mbps and it uses quite alot to do a speed test on ookla. on average between 200 to 300mb per test including upload. if spdmerlin was similar you would go through about 9.6gb to 14.4gb in a 24 hour period. imagine leaving it running all the time imagine how much data that is in a month. if you have a data cap before your ISP charges you extra, then i would not advise running spdmerlin.
 
the obvious explanation i can find is the connmon ping and the spdmerlin test use data. (probably spdmerlin more so) if you do a normal speed test on like lets say ookla for example you are using data to do such. i have my connection dropped down to 250mbps and it uses quite alot to do a speed test on ookla. on average between 200 to 300mb per test including upload. if spdmerlin was similar you would go through about 9.6gb to 14.4gb in a 24 hour period. imagine leaving it running all the time imagine how much data that is in a month. if you have a data cap before your ISP charges you extra, then i would not advise running spdmerlin.

I agree that running constant speed test uses a lot of data. What I would like to see if possible if for the router to be included as a device so you could see how much data it is actually using and to what apps.
 
Last edited:
Just to satisfy my curiosity on how much data is consumed running 48 speed tests daily uses I pulled out my trusty old N66, installed Entware and then SPDmerlin. I then set it up to run on my test bench. The N66's radios were off and nothing was connected to the LAN ports except when I plugged in a cable to check data use.

My speed from my ISP consistently tests at 180/12 Mbps. The N66 with its 600 Mhz processor could only manage an average download speed of 73 Mbps when running a speed test. (In comparison my AC86 averages 173 Mbps).

The average data used daily primarily by SPDmerlin was 5.25 GB. (4.52 GB down - .73 GB up ).

You can draw your own conclusions on how much data would be used with both a faster ISP connection and a router with a faster processor.
 
Just to satisfy my curiosity on how much data is consumed running 48 speed tests daily uses I pulled out my trusty old N66, installed Entware and then SPDmerlin. I then set it up to run on my test bench. The N66's radios were off and nothing was connected to the LAN ports except when I plugged in a cable to check data use.

My speed from my ISP consistently tests at 180/12 Mbps. The N66 with its 600 Mhz processor could only manage an average download speed of 73 Mbps when running a speed test. (In comparison my AC86 averages 173 Mbps).

The average data used daily primarily by SPDmerlin was 5.25 GB. (4.52 GB down - .73 GB up ).

You can draw your own conclusions on how much data would be used with both a faster ISP connection and a router with a faster processor.
Faster ISP and router should not change amount of data, but rather just down/upload it in less time.
 
Faster ISP and router should not change amount of data, but rather just down/upload it in less time.

My understanding is that a speedtest tries to download as much data as it can in ten seconds then repeats the process for the upload. There is a maximum data file size so at some point what you say becomes true but at what point that is I don't know so I didn't try and extrapolate my test results.

In any case even if the use is just five GBs a day for a month could cause a cap issue for heavy data users.
 
Just to satisfy my curiosity on how much data is consumed running 48 speed tests daily uses I pulled out my trusty old N66, installed Entware and then SPDmerlin. I then set it up to run on my test bench. The N66's radios were off and nothing was connected to the LAN ports except when I plugged in a cable to check data use.

My speed from my ISP consistently tests at 180/12 Mbps. The N66 with its 600 Mhz processor could only manage an average download speed of 73 Mbps when running a speed test. (In comparison my AC86 averages 173 Mbps).

The average data used daily primarily by SPDmerlin was 5.25 GB. (4.52 GB down - .73 GB up ).

You can draw your own conclusions on how much data would be used with both a faster ISP connection and a router with a faster processor.
that is not that bad i suppose ~ roughly 140-150gb in a month, my average consumes about 250~260gb, if i disable my QOS it would consume a lot more.

The faster the isp the bigger the file needs to be to determine the isp speed.
 
Just wanted to chime in and thank jackyaz for the prolific work of late. I tried spdmerlin and it worked as expected. It did report upload speeds about half of what it actually is though (verified with several clients on the lan). It was cool to have an extra set of tool graphs to play with, but at the end of the day I uninstalled just because I didn't see the value in consuming so much data, given my connection has always been solid as a rock (have fiber in home, so...). Good to know the tool works, in case my isp ever tries to play games with my service.

Thanks,
Kevin
 
Feature request: Ability to set frequency of tests and possible a blackout period (peak hours). Obviously I can manually delete and readd the cru entry, but that seems less than ideal.

I'd like to run this maybe 2-3 times per day during off-peak hours to collect the data without interrupting regular usage. I have not had good results configuring QoS on my router and when this test runs it can cause some interruptions on the network especially if streaming.
 
Feature request: Ability to set frequency of tests and possible a blackout period (peak hours). Obviously I can manually delete and readd the cru entry, but that seems less than ideal.

I'd like to run this maybe 2-3 times per day during off-peak hours to collect the data without interrupting regular usage. I have not had good results configuring QoS on my router and when this test runs it can cause some interruptions on the network especially if streaming.
It's on the to-do already :)
 
Hey,

I just installed it on my AC87U but the speeds i get are very different to those i get with standard in-browser testing.

Browser: 500+ Mbit/s Download 200+ Mbit/s Upload
Router: 190 Mbit/s Download 25 Mbit/s Upload

Is this due to bottlenecking of the CPU/RAM from the Router?
 
Hey,

I just installed it on my AC87U but the speeds i get are very different to those i get with standard in-browser testing.

Browser: 500+ Mbit/s Download 200+ Mbit/s Upload
Router: 190 Mbit/s Download 25 Mbit/s Upload

Is this due to bottlenecking of the CPU/RAM from the Router?

Yes.
 
Hi,

Just noticed something odd on the Tools page of my RT-AC68U :

Asus-Feedback.png


I have a feedback page titled 'SpeedTest' !

I'm not sure when this happened, but a couple of days ago I did install spdMerlin, but then decided I didn't really need it so uninstalled it (via the menu option) after trying it out briefly.
And then today I updated my router to 384.11 (from 384.11 beta2), so not sure at what point this page appeared like this.

And ideas? Only thing I've tried is a reboot to no avail. It's not causing any issues that I'm aware of, just a bit odd.
 
Me again ....

Solved already, but not sure how I ended up with the above set-up. It seems I either hadn't uninstalled spdMerlin, or the uninstall didn't work, as found that I still had the spdmerlin script in my scripts directory - I'm fairly confident that I did run it.....I think....

Seems odd in either case that I was seeing this feedback page though. Anyway, running the uninstall script (again?) sorted it.

Sorry for the noise....
 
I can not seem to get the graphs to every display anything. other scripts like ntp, ping, diversion, all show graphs but spdmerlin shows the latest ran value on the top but not graphs. Any advice to fix it?
 
I was waiting to see someone report on this one - I've not updated yet as I was mulling over if it is still useful due to the high speed limitations

I would consider dropping it for now. Maybe just one last refresh to show the graphs for those that still find it useful?

You have lots of projects that you can concentrate your time better on, I think.

Not only is the lower specs of our routers skewing results, but having it continuously run is also eating into people's bandwidth per month budgets too.

I had one customer that this would be useful for, but they recently moved to Fibre (not the fastest possible, but many times what their old DSL line was offering).
 
I like this script and the information it provides but wish it was more customizable such as a set schedule (after hours testing) or on demand. My connection is a 120 down 20 up cable internet so it’s nice to make sure I’m getting what I pay for.
 
Ok after reinstalling it i got graphs to finally show.
I also appreciate that you put the pixelserv into the diversion stats.
This speed test would be way better if we can saying to run a test a specified time. 2x an hour is a little much.
 

Similar threads

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top