spdMerlin spdMerlin - v4.x - speedtest and bandwidth monitoring

[Jeffrey Young]

Starting speedtest using Spectrum (Syracuse, NY, United States) for WAN interface

      Server: Spectrum - Syracuse, NY (id: 16978)
         ISP: Spectrum
Idle Latency:    23.03 ms   (jitter: 0.52ms, low: 21.52ms, high: 23.53ms)
    Download:  1133.54 Mbps (data used: 1.4 GB)                                                  
      Upload:    38.97 Mbps (data used: 46.1 MB)                                                  
Packet Loss:     0.0%
Result URL: https://www.speedtest.net/result/14100872244

Speedtest results - Download: 1133.54 Mbps (data used: 1.4 GB) - Upload: 38.97 Mbps (data used: 46.1 MB)

Connection quality - Idle Latency: 23.03 ms (jitter: 0.52ms, low: 21.52ms, high: 23.53ms) - Packet Loss: 0.0%

Retrieving data for WebUI charts


Press enter to continue...

Great, and that is using the built in binary?

 

[visortgw]

Great, and that is using the built in binary?

Yes, it is! dos2unix did the trick!

9.    Toggle between built-in Ookla speedtest and speedtest-cli
      Currently: builtin will be used for speedtests

I also changed version number of my local version number to 4.4.5 (my OCD kicking in!).

 

[Jeffrey Young]

Yes, it is! dos2unix did the trick!

9.    Toggle between built-in Ookla speedtest and speedtest-cli
      Currently: builtin will be used for speedtests

Great, I moidified the instructions. Not sure how to fix pastebin for that. Now to learn how to do a pull request for @Jack Yaz

Just downloaded GitDesktop for Windows. Now to play (and hopfully not screw something up)

 

[Jeffrey Young]

Yes, it is! dos2unix did the trick!

9.    Toggle between built-in Ookla speedtest and speedtest-cli
      Currently: builtin will be used for speedtests

I also changed version number of my local version number to 4.4.5 (my OCD kicking in!).

One more thing, can you check the GUI and make sure the test shows up there?

 

[visortgw]

One more thing, can you check the GUI and make sure the test shows up there?

My auto speed tests are scheduled every 30 minutes (12,42). Results added after downloading your changes switching back to built-in.

 

[chris_d277]

I applied the changes as per post #135, this is the output from my test with the built in binary with 388.1:

Starting speedtest using Telstra (Sydney, Australia) for WAN interface

     Server: Telstra - Sydney (id = 12492)
        ISP: Aussie Broadband
    Latency:     2.08 ms   (0.49 ms jitter)
   Download:   275.81 Mbps (data used: 340.3 MB )
     Upload:    22.51 Mbps (data used: 28.0 MB )
Packet Loss:     0.0%
Result URL: https://www.speedtest.net/result/3%3A17+AM
Error: near line 1: in prepare, near ",": syntax error (1)

Speedtest results - Download: 275.81 Mbps (data used: 340.3 MB ) - Upload: 22.51 Mbps (data used: 28.0 MB )

Connection quality - Latency: 2.08 ms (0.49 ms jitter) - Packet Loss: 0.0%

Retrieving data for WebUI charts

Note the syntax error, and also the webgui graph didn't update either

UPDATE: just performed another run as follows:

Starting speedtest using Telstra (Sydney, Australia) for WAN interface

     Server: Telstra - Sydney (id = 12492)
        ISP: Aussie Broadband
    Latency:     1.99 ms   (0.36 ms jitter)
   Download:   276.76 Mbps (data used: 340.1 MB )
     Upload:    22.55 Mbps (data used: 27.9 MB )
Packet Loss:     0.0%
Result URL: https://www.speedtest.net/result/3%3A46+AM
Parse error near line 1: near ",": syntax error
  rID],[ServerName]) values(1671594413,276.76,22.55,,,0.0,'https://www.speedtest
                                      error here ---^

Speedtest results - Download: 276.76 Mbps (data used: 340.1 MB ) - Upload: 22.55 Mbps (data used: 27.9 MB )

Connection quality - Latency: 1.99 ms (0.36 ms jitter) - Packet Loss: 0.0%

Retrieving data for WebUI charts

So looks like the syntax error is related to the parsing of the result URL, the result URL is also invalid maybe becuase this specific server doesn't have the new low and high latency?

UPDATE2: Switching back to the released v4.4.4, and all works fine with 388.1 with this speedtest server. No syntax errors and charts all updating.

 

[Jeffrey Young]

I applied the changes as per post #135, this is the output from my test with the built in binary with 388.1:

Starting speedtest using Telstra (Sydney, Australia) for WAN interface

     Server: Telstra - Sydney (id = 12492)
        ISP: Aussie Broadband
    Latency:     2.08 ms   (0.49 ms jitter)
   Download:   275.81 Mbps (data used: 340.3 MB )
     Upload:    22.51 Mbps (data used: 28.0 MB )
Packet Loss:     0.0%
Result URL: https://www.speedtest.net/result/3%3A17+AM
Error: near line 1: in prepare, near ",": syntax error (1)

Speedtest results - Download: 275.81 Mbps (data used: 340.3 MB ) - Upload: 22.51 Mbps (data used: 28.0 MB )

Connection quality - Latency: 2.08 ms (0.49 ms jitter) - Packet Loss: 0.0%

Retrieving data for WebUI charts

Note the syntax error, and also the webgui graph didn't update either

UPDATE: just performed another run as follows:

Starting speedtest using Telstra (Sydney, Australia) for WAN interface

     Server: Telstra - Sydney (id = 12492)
        ISP: Aussie Broadband
    Latency:     1.99 ms   (0.36 ms jitter)
   Download:   276.76 Mbps (data used: 340.1 MB )
     Upload:    22.55 Mbps (data used: 27.9 MB )
Packet Loss:     0.0%
Result URL: https://www.speedtest.net/result/3%3A46+AM
Parse error near line 1: near ",": syntax error
  rID],[ServerName]) values(1671594413,276.76,22.55,,,0.0,'https://www.speedtest
                                      error here ---^

Speedtest results - Download: 276.76 Mbps (data used: 340.1 MB ) - Upload: 22.55 Mbps (data used: 27.9 MB )

Connection quality - Latency: 1.99 ms (0.36 ms jitter) - Packet Loss: 0.0%

Retrieving data for WebUI charts

So looks like the syntax error is related to the parsing of the result URL, the result URL is also invalid maybe becuase this specific server doesn't have the new low and high latency?

UPDATE2: Switching back to the released v4.4.4, and all works fine with 388.1 with this speedtest server. No syntax errors and charts all updating.

Thank you for your efforts @chris_d277 in testing. Your input is most valuable. It would seem that some servers to indeed not report buffer bloat while others do. I will have to change my approach to determine how to parse the output. This should also make the script more backwards compatible as well.

I will fix today and post a new test version to try out.

Cheers

 

[Jeffrey Young]

I updated the test spdmerlin script. I was using merlin code base as the test to decide which parsing scheme to use. Thanks to @chris_d277 testing, it would seem that not all servers return buffer bloat data, so I altered the script to check the actual results to see if any buffer bloat data is there in order to decide how to parse the output.

I updated post 135. For those who have already made a backup copy of the production spdmerlin, skip step 1

@visortgw @chris_d277 would you guys be willing to try to the revised script?

 

[visortgw]

I updated the test spdmerlin script. I was using merlin code base as the test to decide which parsing scheme to use. Thanks to @chris_d277 testing, it would seem that not all servers return buffer bloat data, so I altered the script to check the actual results to see if any buffer bloat data is there in order to decide how to parse the output.

I updated post 135. For those who have already made a backup copy of the production spdmerlin, skip step 1

@visortgw @chris_d277 would you guys be willing to try to the revised script?

Give me a few minutes.

 

[visortgw]

I updated the test spdmerlin script. I was using merlin code base as the test to decide which parsing scheme to use. Thanks to @chris_d277 testing, it would seem that not all servers return buffer bloat data, so I altered the script to check the actual results to see if any buffer bloat data is there in order to decide how to parse the output.

I updated post 135. For those who have already made a backup copy of the production spdmerlin, skip step 1

@visortgw @chris_d277 would you guys be willing to try to the revised script?

No errors thrown here.

 

[Jeffrey Young]

No errors thrown here.

Great. Thanks for the help

 

[visortgw]

It's the least that I could do. You did all of the heavy lifting!

 

[miniterror]

I updated the test spdmerlin script. I was using merlin code base as the test to decide which parsing scheme to use. Thanks to @chris_d277 testing, it would seem that not all servers return buffer bloat data, so I altered the script to check the actual results to see if any buffer bloat data is there in order to decide how to parse the output.

I updated post 135. For those who have already made a backup copy of the production spdmerlin, skip step 1

@visortgw @chris_d277 would you guys be willing to try to the revised script?

Great job men, although im still running old version of Merlin its appriciated you put your time/effort into this.

 

[Jeffrey Young]

@Jack Yaz

I issued a Pull Request to your devopment branch. Or, at least I think I did. First time doing this. Can you have a look to make sure it went through.

Cheers

 

[Jack Yaz]

@Jack Yaz

I issued a Pull Request to your devopment branch. Or, at least I think I did. First time doing this. Can you have a look to make sure it went through.

Cheers

all merged and ready for testing on develop branch

spdmerlin develop

 

[Jeffrey Young]

Perfect, speedtest run and GUI updated on 386.7_2 OK.

 

[Stephen Harrington]

Also reporting success?
Updated to 4.4.5 dev on my RT-AX86U running 386.7_2, confirmed speed tests and GUI were still running/updating fine (internal binary).
I then did a "dirty" upgrade to 388.1 and "early days" but confirmed everything still seems to be ok.

 

[chris_d277]

Everything seems to work fine for me with v4.4.5 dev on 388.1, thanks everyone

 

[aex.perez]

An Observation, just curious,

Under 4.4.4 and now 4.4.5, the amounts of data uploaded/downloaded is symetrical and more/larger and the same amounts on every run using the "Built-In" binary vs. using the "External" binary that has the uploaded/downloaded data amounts asymetrical, less/lower and different amounts on every run?

Ex. Uploaded/Downloaded when set to "Built-In" is consistently 1.2MB for both on every run. Using the "External" binary, under the same condititions, the Uploaded/Downloaded data differs for each and on every run and more often less than, sometimes much less than 1.2MB for either.

In both cases achieving the same results, always WAN interface using preferred server but same observation if using a different preferred server - haven't used Auto Select in awhile.

 

[aex.perez]

An Observation, just curious,

Under 4.4.4 and now 4.4.5, the amounts of data uploaded/downloaded is symetrical and more/larger and the same amounts on every run using the "Built-In" binary vs. using the "External" binary that has the uploaded/downloaded data amounts asymetrical, less/lower and different amounts on every run?

Ex. Uploaded/Downloaded when set to "Built-In" is consistently 1.2MB for both on every run. Using the "External" binary, under the same condititions, the Uploaded/Downloaded data differs for each and on every run and more often less than, sometimes much less than 1.2MB for either.

In both cases achieving the same results, always WAN interface using preferred server but same observation if using a different preferred server - haven't used Auto Select in awhile.

Some examples;

Speedtest result for WAN - External
Bandwidth - Download: 928.48 Mbps (data used: 821.8 MB ) - Upload: 907.10 Mbps (data used: 844.6 MB )
Quality - Latency: 3.60 ms (0.14 ms jitter) - Packet Loss: Not available.
Result URL: https://www.speedtest.net/result/c/60aa7d29-0480-4bc1-a367-0a3b4aa87c0b

Speedtest result for WAN - External
Bandwidth - Download: 928.21 Mbps (data used: 476.5 MB ) - Upload: 921.23 Mbps (data used: 958.0 MB )
Quality - Latency: 4.23 ms (0.06 ms jitter) - Packet Loss: 0.0%
Result URL: https://www.speedtest.net/result/c/de7e7e56-1c7c-464e-b156-14575fb9a723


Speedtest result for WAN - Built In
Bandwidth - Download: 887.53 Mbps (data used: 1.1 GB; 3 additional servers) - Upload: 875.90 Mbps (data used: 1.1 GB )
Quality - Latency: 4.05 ms (0.21 ms jitter) - Packet Loss: Not available.
Result URL: https://www.speedtest.net/result/14106778883

Speedtest result for WAN - Built In
Bandwidth - Download: 931.43 Mbps (data used: 1.2 GB ) - Upload: 923.75 Mbps (data used: 1.1 GB )
Quality - Latency: 4.16 ms (0.26 ms jitter) - Packet Loss: 0.0%
Result URL: https://www.speedtest.net/result/14106784673

Also noticed what could be a bug, but not sure if introduced in 4.4.5 or a prior release.
Changed the amount of days to display/keep to their mximums (not sure if any other values create the issue)

Once I did that, the bandwidth test would not run/give results regardless of which binary (Built-In or External) I used, even if I changed the values back to what they originally were - the test would still fail to run. I had to reselect the preferred server again to fix it.