spdMerlin spdMerlin - v4.x - speedtest and bandwidth monitoring

[Viktor Jaep]

If someone wants to buy me an Asus router....I kid of course.

I will need your credit card number to put a hold on this GT-AXE16000, so if you can please PM that over to me, I'll have this shipped right out to you!

can someone run the speedtest binaries (without spdmerlin) and post a copy of the results output each cough out?

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --format=csv --interface=eth0 --server-id=17360

"CenturyLink - Phoenix, AZ","17360","58.92","1.824","N/A","41920330","4754289","602203176","63773915","https://www.speedtest.net/result/c/494c608d-bddf-4c92-8ba2-6a3255115fa9","1"

 

[Matthew Patrick]

I just noticed that speedtest now tests bufferbloat on download and upload tests. Is it possible for spdmerlin to put that into the tables? I have tried downloading and running the prebuilt Speedtest binary for aarch64/arm64 and it does work. Is it something that's gonna be implemented later on? Thank you

 

[Jack Yaz]

I just noticed that speedtest now tests bufferbloat on download and upload tests. Is it possible for spdmerlin to put that into the tables? I have tried downloading and running the prebuilt Speedtest binary for aarch64/arm64 and it does work. Is it something that's gonna be implemented later on? Thank you

its probably the bufferbloat output messing with the output to be honest - does the external binary option against the same server include it?

 

[Viktor Jaep]

its probably the bufferbloat output messing with the output to be honest - does the external binary option against the same server include it?

Honestly, I'm not seeing bufferbloat when running the binary... I don't see it on their site either. Where are you finding this, @Matthew Patrick?

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --interface=eth0 --server-id=17360

   Speedtest by Ookla

     Server: CenturyLink - Phoenix, AZ (id = 17360)
        ISP: Comcast Cable
    Latency:    58.25 ms   (2.19 ms jitter)
   Download:   326.68 Mbps (data used: 568.4 MB )
     Upload:    37.66 Mbps (data used: 41.1 MB )
Packet Loss: Not available.
 Result URL: https://www.speedtest.net/result/c/26e3fb0b-1dd0-4467-a54b-b3dd5ba46fba

 

[SomeWhereOverTheRainBow]

Honestly, I'm not seeing bufferbloat when running the binary... I don't see it on their site either. Where are you finding this, @Matthew Patrick?

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --interface=eth0 --server-id=17360

   Speedtest by Ookla

     Server: CenturyLink - Phoenix, AZ (id = 17360)
        ISP: Comcast Cable
    Latency:    58.25 ms   (2.19 ms jitter)
   Download:   326.68 Mbps (data used: 568.4 MB )
     Upload:    37.66 Mbps (data used: 41.1 MB )
Packet Loss: Not available.
Result URL: https://www.speedtest.net/result/c/26e3fb0b-1dd0-4467-a54b-b3dd5ba46fba

here is what mine outputs....

SWOTR@RT-AX88U-C7C0:/usr/sbin# ./ookla
[2022-12-07 00:52:01.715] [error] ConfigurationError - Could not retrieve or read configuration (Configuration)
{"type":"log","timestamp":"2022-12-07T05:52:01Z","message":"Configuration - Could not retrieve or read configuration (ConfigurationError)","level":"error"}

 

[Matthew Patrick]

its probably the bufferbloat output messing with the output to be honest - does the external binary option against the same server include it?

Honestly, I'm not seeing bufferbloat when running the binary... I don't see it on their site either. Where are you finding this, @Matthew Patrick?

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --interface=eth0 --server-id=17360

   Speedtest by Ookla

     Server: CenturyLink - Phoenix, AZ (id = 17360)
        ISP: Comcast Cable
    Latency:    58.25 ms   (2.19 ms jitter)
   Download:   326.68 Mbps (data used: 568.4 MB )
     Upload:    37.66 Mbps (data used: 41.1 MB )
Packet Loss: Not available.
Result URL: https://www.speedtest.net/result/c/26e3fb0b-1dd0-4467-a54b-b3dd5ba46fba

I meant that since a few months ago. Speedtest.net now checks bufferbloat as in the ping times when saturating download and upload. It used to only check the ping on idle. Which doesn't help in determining bufferbloat under load.. the new site and mobile apps now checks the ping times on idle, download and upload.. the CLI is now doing the same thing too but it seems like the binary in spdMerlin is still using the old one.

That's why I asked if it's gonna be supported on spdMerlin in the future

For the binary I used. I downloaded and tried the latest prebuilt Speedtest binary from speedtest.net . The latest one now tests ping under load for both download and upload tests

 

[miniterror]

I think he means the lower part of the lower part of the screenshot.
If thats the case that would mean the binary in the newer Rmerlin build has updated to a version that supports this.
On the 386.7_2 is there a way i can check the used built-in binary version?
Then it could be compared to someone running the 388.1 version.

 

[Viktor Jaep]

I meant that since a few months ago. Speedtest.net now checks bufferbloat as in the ping times when saturating download and upload. It used to only check the ping on idle. Which doesn't help in determining bufferbloat under load.. the new site and mobile apps now checks the ping times on idle, download and upload.. the CLI is now doing the same thing too but it seems like the binary in spdMerlin is still using the old one.

That's why I asked if it's gonna be supported on spdMerlin in the future

For the binary I used. I downloaded and tried the latest prebuilt Speedtest binary from speedtest.net . The latest one now tests ping under load for both download and upload tests

Sure enough! It looks like ookla came out with a long-overdue CLI update... from v1.1.1 to v1.2.0. Here are the new outputs:

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --format=csv --interface=eth0 --server-id=10051
"Comcast - Denver, CO","10051","46.7325","1.715","0","40366508","4887021","373309128","49648696","https://www.speedtest.net/result/c/953a92ce-b3b6-47a7-be14-bbc8adf078ff","1","53.6259","24.2515","41.317","371.042","54.2089","7.28236","42.223","245.64","44.563","48.266"

It looks like they tacked a bunch of new values on behind the URL as to not break existing code. @Jack Yaz, do you have any access to documentation on the CLI that might be able to break down what all these values mean? Here's the human-readable view:

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --interface=eth0 --server-id=10051

   Speedtest by Ookla

      Server: Comcast - Denver, CO (id: 10051)
         ISP: Comcast Cable
Idle Latency:    46.12 ms   (jitter: 1.63ms, low: 44.42ms, high: 47.71ms)
    Download:   344.06 Mbps (data used: 361.4 MB)                                                 
                 57.20 ms   (jitter: 28.98ms, low: 44.39ms, high: 383.44ms)
      Upload:    39.58 Mbps (data used: 44.1 MB)
                 55.34 ms   (jitter: 7.87ms, low: 40.89ms, high: 261.81ms)
 Packet Loss: Not available.
  Result URL: https://www.speedtest.net/result/c/ce36ead8-0cdf-4d74-8b50-2cced7c0ffdd

 

[Viktor Jaep]

@Jack Yaz ... looks there's a switch that explains what all the CSV fields mean. Whew. I'm going to start working on getting these implemented in RTRMON.

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --output-header --format=csv --interface=eth0 --server-id=10051
"server name","server id","idle latency","idle jitter","packet loss","download","upload","download bytes","upload bytes","share url","download server count","download latency","download latency jitter","download latency low","download latency high","upload latency","upload latency jitter","upload latency low","upload latency high","idle latency low","idle latency high"
"Comcast - Denver, CO","10051","45.7506","0.793","0","38949988","4916662","345414856","50199672","https://www.speedtest.net/result/c/0a07dc74-25a5-424b-941e-cfd724791670","1","54.5774","21.9203","42.091","366.561","55.0167","7.97647","42.509","232.395","44.537","46.953"

 

[Mogsy]

Is this why my buffer size was huge in scMerlin? I was doing speedtest on their web and app. No spdMerlin installed tho

 

[SomeWhereOverTheRainBow]

@Jack Yaz ... looks there's a switch that explains what all the CSV fields mean. Whew. I'm going to start working on getting these implemented in RTRMON.

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --output-header --format=csv --interface=eth0 --server-id=10051
"server name","server id","idle latency","idle jitter","packet loss","download","upload","download bytes","upload bytes","share url","download server count","download latency","download latency jitter","download latency low","download latency high","upload latency","upload latency jitter","upload latency low","upload latency high","idle latency low","idle latency high"
"Comcast - Denver, CO","10051","45.7506","0.793","0","38949988","4916662","345414856","50199672","https://www.speedtest.net/result/c/0a07dc74-25a5-424b-941e-cfd724791670","1","54.5774","21.9203","42.091","366.561","55.0167","7.97647","42.509","232.395","44.537","46.953"

....

 

[Viktor Jaep]

Added the new Ookla speedtest CLI features into RTRMON...

 

[SomeWhereOverTheRainBow]

Added the new Ookla speedtest CLI features into RTRMON...

View attachment 46109

RTRMON featuring all the speed tests you know and love:-Coming to a terminal near you this holiday season! Get Ready!

Potential additional holiday terminal goodies:-
MWANMON
NTPFIXMON
HALTMON
ADTRAKMON
WGMADMON

@Viktor Jaep keep up the Great work!

 

[Jack Yaz]

@Jack Yaz ... looks there's a switch that explains what all the CSV fields mean. Whew. I'm going to start working on getting these implemented in RTRMON.

ViktorJp@RT-AC86U-BE10:/jffs/addons/rtrmon.d# ./speedtest --output-header --format=csv --interface=eth0 --server-id=10051
"server name","server id","idle latency","idle jitter","packet loss","download","upload","download bytes","upload bytes","share url","download server count","download latency","download latency jitter","download latency low","download latency high","upload latency","upload latency jitter","upload latency low","upload latency high","idle latency low","idle latency high"
"Comcast - Denver, CO","10051","45.7506","0.793","0","38949988","4916662","345414856","50199672","https://www.speedtest.net/result/c/0a07dc74-25a5-424b-941e-cfd724791670","1","54.5774","21.9203","42.091","366.561","55.0167","7.97647","42.509","232.395","44.537","46.953"

I don't use csv output format as I display the interactive test both in CLI and WebUI, and parse the results. Not ideal but it looks shiny

 

[Viktor Jaep]

RTRMON featuring all the speed tests you know and love:-Coming to a terminal near you this holiday season! Get Ready!

Potential additional holiday terminal goodies:-
MWANMON
NTPFIXMON
HALTMON
ADTRAKMON
WGMADMON

@Viktor Jaep keep up the Great work!

Last but not least...

WTFMON

 

[Rassal]

I am also getting this same problem with spdMerlin... it runs, but no charts...

Is there any solution?

 

[miniterror]

I am also getting this same problem with spdMerlin... it runs, but no charts...

Is there any solution?

Built-in version has been updated and is giving more info then what it used to give.
Im thinking this is breaking the import of the data into the file that creates the graphs.
I'm not sure if the creator will update this tool so it can handle it again as he isnt using a Asus anymore.

Side question (feature request) if the the creator decides to update this tool.
Would it be possible if a test fails to generate it as 0 in the logfiles, that way we can see a dip in the charts and easily spot failed attempts instead of trying to find missing dots.


EDIT
Looking in the config file it seems like its pulling the Built-in from /usr/sbin/ookla.
One could try to replace that file with the old version from ookla and hope it starts working again.
Dont forget to give the file 0755 rights, it has those now too.
For people wanting to test i have extracted the old ookla file from my 386.7_2 installation.
You can download it here: ookla

 

[gjf]

Any idea why running ookla speedtest shows such error at the end (however test works ok)? Also looks like Result URL is wrong...

UPD: Solved - thanks to miniterror findings: it really should be toggled from built-in speedtest to speedtest-cli - in this case both URL and charts problems are getting solved.

 

[Jeffrey Young]

Built-in version has been updated and is giving more info then what it used to give.
Im thinking this is breaking the import of the data into the file that creates the graphs.
I'm not sure if the creator will update this tool so it can handle it again as he isnt using a Asus anymore.

Side question (feature request) if the the creator decides to update this tool.
Would it be possible if a test fails to generate it as 0 in the logfiles, that way we can see a dip in the charts and easily spot failed attempts instead of trying to find missing dots.


EDIT
Looking in the config file it seems like its pulling the Built-in from /usr/sbin/ookla.
One could try to replace that file with the old version from ookla and hope it starts working again.
Dont forget to give the file 0755 rights, it has those now too.
For people wanting to test i have extracted the old ookla file from my 386.7_2 installation.
You can download it here: ookla

Did this start with the 388 code base?

 

[miniterror]

Did this start with the 388 code base?

Yep, that version seems to have the new speedtest binary.