spdMerlin The future of spdMerlin

[Wrkdbf_Guy]

Maybe I'm blind, but where did ASUS tuck the Ookla Speedtest in firmware 386 GUI? I'd like to compare it to spdMerlin, but can't find it. (ASUS documentation is lacking and non-existent in many regards).

Bill

 

[francovilar]

Maybe I'm blind, but where did ASUS tuck the Ookla Speedtest in firmware 386 GUI? I'd like to compare it to Yazfi, but can't find it. (ASUS documentation is lacking and non-existent in many regards).

Bill

Adaptive qos - internet speed

 

[Wrkdbf_Guy]

Adaptive qos - internet speed

Thanks! That's seems like an odd place to put it. If one isn't interested in Adaptive QoS, like me, they'd never think to go there. Tools or Network Tools would seem to make more sense, but whatever. Thanks again.

 

[iTyPsIDg]

currently only works with adaptive qos (but doesn't actually commit to nvram yet). i need to check with @dave14305 on the best way to trigger flexqos if installed after changing bandwidth values

does cake-qos store bandwidth values in a config file, or provide a command line hook?

I don't think anyone ever answered this question. It's stored in /jffs/addons/cake-qos/cake-qos.cfg

I'm not positive, but you may also be able to use command-line options to set these values. command-line options to set these values

 

[vango44]

This is what I get with the Asus test:

vs spdMerlin:

So hopefully a continued addon. Thanks!

 

[miniterror]

Yeah i really like this too.
I was complaining with my VPN provide that the soeeds were not good (only getting arround 60/65Mbit while i have a 250Mbit connection.
Showing the graphs from the VPN client vs WAN i could show its not a unstable ISP connection.
They still didnt solve the problem but atleast got rid of some discussion.
Also really like to see my own stability.
Fun fact, while gaming i always notice it as my ping shoots up for a few seconds ingame as expected and its now a good measure for me to see if all scripts started after the nightly reboot.
So im really pleased with it, my hats of to you for making this

 

[archiel]

I have been using this tool for a while now and really like it. I do keep noticing one issue and not sure if it is just my setup

spdMerlin is set to run on WAN and VPN5 interfaces twice an hour and every so often the test will not run. Once it fails to run then it will not try to run at the next due time, and this applies to both the WAN and VPN interfaces. If I then run a manual test and it is successful, then the autotests will restart on both interfaces.

It does not make a difference whether I am using auto-selected or a preferred server, simply that the test fails, this could be due to an issue with the test server or, as happened yesterday, issues with my local internet connection.

I have been using spdMerlin since at least 384.18 and have done a full WPS reset / manual rebuild on each upgrade, so I not not think 'residual config clutter' should be the issue.

 

[Jack Yaz]

I have been using this tool for a while now and really like it. I do keep noticing one issue and not sure if it is just my setup

spdMerlin is set to run on WAN and VPN5 interfaces twice an hour and every so often the test will not run. Once it fails to run then it will not try to run at the next due time, and this applies to both the WAN and VPN interfaces. If I then run a manual test and it is successful, then the autotests will restart on both interfaces.

It does not make a difference whether I am using auto-selected or a preferred server, simply that the test fails, this could be due to an issue with the test server or, as happened yesterday, issues with my local internet connection.

I have been using spdMerlin since at least 384.18 and have done a full WPS reset / manual rebuild on each upgrade, so I not not think 'residual config clutter' should be the issue.

do you see any errors reported in system log? after the next fail, please try running:

cru l

and check if there is an entry for spdMerlin

 

[archiel]

do you see any errors reported in system log? after the next fail, please try running:

cru l

and check if there is an entry for spdMerlin

There is nothing obvious in the logs, I can see where the script last ran and various activity after the internet went down and I restarted the wi-fi via scMerlin, but everything was up and running with 27 minutes, it is just that the next run that was due (after 30minutes) didn't happen.

I have made a note to check cru l, after the next time it occurs, but this could be anytime, it has so far happened about 5 times in the 7/8 months. I did check cru l now and it looks normal, contents below.

0 * * * * /jffs/addons/cake-qos/cake-qos checkrun #CakeQOS-Merlin#
*/2 * * * * /etc/openvpn/server1/vpn-watchdog1.sh #CheckVPNServer1#
25 16 * * * sh /jffs/scripts/firewall banmalware #Skynet_banmalware#
20 1 * * Mon sh /jffs/scripts/firewall update #Skynet_autoupdate#
0 * * * * sh /jffs/scripts/firewall save #Skynet_save#
56 */12 * * * sh /jffs/scripts/firewall debug genstats #Skynet_genstats#
12,42 0-23 * * * /jffs/scripts/spdmerlin generate #spdMerlin#
00 2 * * Sat sh /opt/share/diversion/file/update-bl.div reset #Diversion_UpdateBL#
20 5 * * * sh /opt/share/diversion/file/rotate-logs.div #Diversion_RotateLogs#
20 17 * * * diversion count_ads count #Diversion_CountAds#
1 0 * * * /opt/bin/find /opt/var/lib/unbound/unbound.log -size +10M -exec rm -f {} \; #unboundLOG#

 

[JT Strickland]

Mine doesn't always work, either, but it's a lot more often than once a month, more like a few times a week. This is a piece of the log where it did not have results:

Feb 27 08:12:30 RT-AC86U-8F38 spdMerlin: Connection quality - Latency: 54.07 ms (0.43 ms jitter) - Packet Loss: 0.0%
Feb 27 08:42:00 RT-AC86U-8F38 spdMerlin: Starting speedtest using auto-selected server for WAN interface
Feb 27 08:42:11 RT-AC86U-8F38 spdMerlin: Speedtest results - -
Feb 27 08:42:11 RT-AC86U-8F38 spdMerlin: Connection quality - -
Feb 27 08:53:51 RT-AC86U-8F38 kernel: br0: received packet on eth6 with own address as source addre

But it will run on time and eventually get results, often the next time.

Edit: I just had one without any data about 8 minutes ago, and this is my output:

ASUSWRT-Merlin RT-AC86U 384.19_0 Fri Aug 14 19:19:51 UTC 2020
jtstrickland@RT-AC86U-8F38:/tmp/home/root# cru l
5 0 * * * /opt/sbin/logrotate /opt/etc/logrotate.conf >> /opt/tmp/logrotate.daily 2>&1 #logrotate#
00 4 * * Sun sh /opt/share/diversion/file/update-bl.div reset #Diversion_UpdateBL#
00 4 * * Wed sh /opt/share/diversion/file/update-bl.div #Diversion_UpdateBLTwo#
20 5 * * * sh /opt/share/diversion/file/rotate-logs.div #Diversion_RotateLogs#
20 17 * * * diversion count_ads count #Diversion_CountAds#
30 3 * * Sun sh /opt/share/diversion/file/stats.div #Diversion_WeeklyStats#
50 3 * * Sun sh /opt/share/diversion/file/backup.div #Diversion_AutoBackup#
10 3 * * Sun sh /opt/share/diversion/file/update.div #Diversion_UpdateCheck#
*/5 0-23 * * * /jffs/scripts/connmon generate #connmon#
25 21 * * * sh /jffs/scripts/firewall banmalware #Skynet_banmalware#
23 1 * * Mon sh /jffs/scripts/firewall update #Skynet_autoupdate#
0 * * * * sh /jffs/scripts/firewall save #Skynet_save#
*/10 * * * * /jffs/scripts/ntpmerlin generate #ntpMerlin#
12,42 0-23 * * * /jffs/scripts/spdmerlin generate #spdMerlin#
0 * * * * /jffs/scripts/uiDivStats generate #uiDivStats_generate#
3 0 * * * /jffs/scripts/uiDivStats trimdb #uiDivStats_trim#
* * * * * /jffs/scripts/uiDivStats querylog #uiDivStats_querylog#
4,9,14,19,24,29,34,39,44,49,54,59 * * * * /jffs/scripts/uiDivStats flushtodb #uiDivStats_flushtodb#
55 */12 * * * sh /jffs/scripts/firewall debug genstats #Skynet_genstats#
jtstrickland@RT-AC86U-8F38:/tmp/home/root#

 

[archiel]

There is nothing obvious in the logs, I can see where the script last ran and various activity after the internet went down and I restarted the wi-fi via scMerlin, but everything was up and running with 27 minutes, it is just that the next run that was due (after 30minutes) didn't happen.

I have made a note to check cru l, after the next time it occurs, but this could be anytime, it has so far happened about 5 times in the 7/8 months. I did check cru l now and it looks normal, contents below.

0 * * * * /jffs/addons/cake-qos/cake-qos checkrun #CakeQOS-Merlin#
*/2 * * * * /etc/openvpn/server1/vpn-watchdog1.sh #CheckVPNServer1#
25 16 * * * sh /jffs/scripts/firewall banmalware #Skynet_banmalware#
20 1 * * Mon sh /jffs/scripts/firewall update #Skynet_autoupdate#
0 * * * * sh /jffs/scripts/firewall save #Skynet_save#
56 */12 * * * sh /jffs/scripts/firewall debug genstats #Skynet_genstats#
12,42 0-23 * * * /jffs/scripts/spdmerlin generate #spdMerlin#
00 2 * * Sat sh /opt/share/diversion/file/update-bl.div reset #Diversion_UpdateBL#
20 5 * * * sh /opt/share/diversion/file/rotate-logs.div #Diversion_RotateLogs#
20 17 * * * diversion count_ads count #Diversion_CountAds#
1 0 * * * /opt/bin/find /opt/var/lib/unbound/unbound.log -size +10M -exec rm -f {} \; #unboundLOG#

It stopped again, though this time it was the speed-test on the VPN interface that failed.

Checked cru l and it was exactly the same as before

The script should have run at 19:12, then 19:42, so I went through the logs to see anything at that time (not contiguous as I use Scribe to split into bite-size chunks)

at 19:12

Mar  4 19:12:00 RT-AX88U-5050 spdMerlin: Starting speedtest using UK Dedicated Servers (London, United Kingdom) for WAN interface
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Speedtest results - Download: 53.68 Mbps (data used: 98.6 MB) - Upload: 18.10 Mbps (data used: 10.4 MB)
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Connection quality - Latency: 5.52 ms (0.99 ms jitter) - Packet Loss: 0.0%
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Starting speedtest using using auto-selected server for VPNC5 interface

followed by
Mar  4 19:12:44 RT-AX88U-5050 kernel: br0: port 5(eth5) neighbor 7f00.d4:da:cd:4b:13:c9 lost

With nothing further referencing spdMerlin until I ran a manual speed test at 19:44 (I could see nothing of note at 19:42)

Mar  4 19:42:00 RT-AX88U-5050 spdMerlin: Stale lock file found (>600 seconds old) - purging lock
Mar  4 19:42:00 RT-AX88U-5050 spdMerlin: Starting speedtest using UK Dedicated Servers (London, United Kingdom) for WAN interface

After which the missing 19:12 test results appear under last 10 speedtest results (it was missing) and the auto test for 19:42 ran (both interfaces) at 19:44

 

[Jack Yaz]

It stopped again, though this time it was the speed-test on the VPN interface that failed.

Checked cru l and it was exactly the same as before

The script should have run at 19:12, then 19:42, so I went through the logs to see anything at that time (not contiguous as I use Scribe to split into bite-size chunks)

at 19:12

Mar  4 19:12:00 RT-AX88U-5050 spdMerlin: Starting speedtest using UK Dedicated Servers (London, United Kingdom) for WAN interface
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Speedtest results - Download: 53.68 Mbps (data used: 98.6 MB) - Upload: 18.10 Mbps (data used: 10.4 MB)
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Connection quality - Latency: 5.52 ms (0.99 ms jitter) - Packet Loss: 0.0%
Mar  4 19:12:24 RT-AX88U-5050 spdMerlin: Starting speedtest using using auto-selected server for VPNC5 interface

followed by
Mar  4 19:12:44 RT-AX88U-5050 kernel: br0: port 5(eth5) neighbor 7f00.d4:da:cd:4b:13:c9 lost

With nothing further referencing spdMerlin until I ran a manual speed test at 19:44 (I could see nothing of note at 19:42)

Mar  4 19:42:00 RT-AX88U-5050 spdMerlin: Stale lock file found (>600 seconds old) - purging lock
Mar  4 19:42:00 RT-AX88U-5050 spdMerlin: Starting speedtest using UK Dedicated Servers (London, United Kingdom) for WAN interface

After which the missing 19:12 test results appear under last 10 speedtest results (it was missing) and the auto test for 19:42 ran (both interfaces) at 19:44

if its never posting the result, it sounds like something is hanging prior to completion. I'll put some verbose logging on develop and lets see if we can find where

 

[Jack Yaz]

if its never posting the result, it sounds like something is hanging prior to completion. I'll put some verbose logging on develop and lets see if we can find where

please run

spdmerlin develop

and leave things to run as normal. next time you notice a failure, please check syslog as I have added some extra messaging

 

[archiel]

please run

spdmerlin develop

and leave things to run as normal. next time you notice a failure, please check syslog as I have added some extra messaging

Installed and I will post when something happens (or rather doesn't happen). Thanks.

 

[Jack Yaz]

by sheer coincidence, one of my VPN tunnels is flaking out and causing spdmerlin to fail in weird ways. develop branch has some changes i'd like others to test.

Changes:

• 2 minute watchdog to prevent speedtests hanging
• Purge any stale speedtest processes before starting another
• Check for the word "FAILED" anywhere in the output and don't attempt to parse data

 

[MvW]

I have a question about SpdMerlin. I've just installed 382 beta 1 and enabled Cake QoS. I have little to no experience with QoS in general but after reading about the advantages of Cake, I figured I might as well give it a try. In the notes it says that leaving bandwith on Automatic will tell Cake basically it has unlimited bandwith.

When I manually enter around 95% of measured up- and downstream, as I have done now, will SpdMerlin still be able to automatically adjust these values? And will that be visible (are the values visually being changed in the manual bandwith values) or does it happen behind the scenes?

 

[ttgapers]

@Jack Yaz - does the native speedtest require accepting anything on the Privacy tab like Ookla/Trend? If so, I vote to keep spdMerlin going for the privacy conscious.

 

[Jack Yaz]

I have a question about SpdMerlin. I've just installed 382 beta 1 and enabled Cake QoS. I have little to no experience with QoS in general but after reading about the advantages of Cake, I figured I might as well give it a try. In the notes it says that leaving bandwith on Automatic will tell Cake basically it has unlimited bandwith.

When I manually enter around 95% of measured up- and downstream, as I have done now, will SpdMerlin still be able to automatically adjust these values? And will that be visible (are the values visually being changed in the manual bandwith values) or does it happen behind the scenes?

currently the AutoBW feature only adjusts Adaptive QoS. If Cake f/w version reads from the same nvram variables then in theory it'll work with no changes from me

 

[MvW]

@Jack Yaz - does the native speedtest require accepting anything on the Privacy tab like Ookla/Trend? If so, I vote to keep spdMerlin going for the privacy conscious.

From the Ookla Speedtest page:

This service is provided by Ookla®. It detects the download and upload speed from your router to the Internet. This test takes approximately one minute to complete. By using this service, you agree to Ookla’s Privacy Policy.

You don’t even have to tick a box, by using it you accept Ooklas TOS.

 

[MvW]

currently the AutoBW feature only adjusts Adaptive QoS. If Cake f/w version reads from the same nvram variables then in theory it'll work with no changes from me

I’ve set the bandwidth setting to manual and see the values changing automagically after a speedtest has run, so it would it actually surprise me if these values are not being used as it’s the only place to configure them in the WebUi, as far as am I’m aware. @RMerlin can you confirm the auto-adjusted values are used for Cake-Qos as well? They are being adjusted in the WebUi, which seems to confirm it are the same variables?