[Fork] Asuswrt-Merlin 374.43 LTS releases (Archive)

[Deleted member 27741]

This version of the firmware packs a lot of improvements- a couple things I've noticed-
OpenVPN update has resolved an issue with tls-auth in my logs.
Dropbear update has significantly sped up putty and plink usage.
I can actually see the cursor in the areas of the firmware that have it- nice for typing especially in vpn options.
Surely there are improvements behind the scenes as well (dns-masq comes to mind).
John, do you use the traffic monitor? Some tweaks that I think would greatly improve the firmware there are-
1) Reception and transmission- same color for all. Unless there is a reason for the inversion? Maybe I'm missing something.
2) Real time- why does the graph not populate the last ten minutes? We know the data is coming from the right and you can see time on the graph when you mouse over. It would be nice not to wait for several minutes to get an idea what is going on in real time. It there a technical reason for blanking the graph?

 

[john9527]

1) Reception and transmission- same color for all. Unless there is a reason for the inversion? Maybe I'm missing something.

Speculating here.....transmission from a client is really reception by the internet and vice versa. So switching between the internet and one of the other tabs would cause the colors to change....and the average user would be asking 'why if I'm doing a download test is the internet tab showing upload'. This way the colors line up between the clients and internet.

2) Real time- why does the graph not populate the last ten minutes? We know the data is coming from the right and you can see time on the graph when you mouse over. It would be nice not to wait for several minutes to get an idea what is going on in real time. It there a technical reason for blanking the graph?

I'm not sure I understand, for me realtime is realtime......if I run a download test for example, it immediately shows up.

 

[Deleted member 27741]

1) That makes sense. I see I was missing something.

2) I mean the real time graph comes up empty then populates from the right. If it were populated with data (from the past 10 minutes or so) AND the data came from the right you still get all the real time and the last ten minutes. I often find myself looking at the real time graph and wishing the last ten minutes were already there. I am assuming data is there to be had, maybe that would be a sweet option? Populate real-time graph with past to minutes yes no?

 

[john9527]

2) I mean the real time graph comes up empty then populates from the right. If it were populated with data (from the past 10 minutes or so) AND the data came from the right

Ahhh....it was me that was missing something Now I understand.
I doubt that they actually collect that level of detail on an ongoing basis, more likely update a fixed set of counters to be able to create the last 24 hour view (speculation). This would keep the data a fixed size and minimize the resources used. But, I agree with your concept that having the last x minutes of detail immediately available would definitely be nice. I like a challenge....but no promises.

 

[Deleted member 27741]

It WOULD be nice, wouldnt it!?

For some reason I made a guess that they were collecting that data constantly on an ongoing basis. At the very least, I will be terribly interested in seeing what you find out. There is no way to turn off traffic monitoring, is there?

 

[daicon]

Hello,

I managed to run a usb 4g Huawei e3372s No- HiLink with Ac68u , Firmware Merlin .

But no way connected to 4g , 3g alone . It is normal?
I think that connects 3G mode that did not reach a higher speed to 18 MB.
In contrast with HiLink so I reach speeds of 60MB .

Do you know what can happen?

 

[Builder71]

... I say you may 'want' to, because most security guidelines are actually recommending a 2048bit key now.

Thx, using a 2048bit key now.
All works as expected.

 

[john9527]

Hello,

I managed to run a usb 4g Huawei e3372s No- HiLink with Ac68u , Firmware Merlin .

But no way connected to 4g , 3g alone . It is normal?
I think that connects 3G mode that did not reach a higher speed to 18 MB.
In contrast with HiLink so I reach speeds of 60MB .

Do you know what can happen?

I have no personal experience with USB modems, but my sense in following the forums is that the modem support regularly changes...some modems would work on a particular release, then fail on the next release. I thinks it's a combination of setup strings and usb drivers, and many times if the modem does work, it won't connect at it's maximum speed.

I believe that Merlin just picks up the latest modem support from Asus in his builds since he does not have any capacity to test usb modems and even if he did, each one is unique.

As far as this fork, it has the modem support that was in place over a year ago, nothing has been updated there. If you want to use this fork, and find the modem works on a later release, I can look at what I can do, but I can't guarantee I'll be able to fix it.

 

[RMerlin]

As far as this fork, it has the modem support that was in place over a year ago, nothing has been updated there. If you want to use this fork, and find the modem works on a later release, I can look at what I can do, but I can't guarantee I'll be able to fix it.

Actually quite a fair amount of work would be required to backport the current USB modem support. It's unfortunately not just a matter of upgrading usb-modeswitch (which does help things out), but also the various rom/*.sh and rc/usb*.c (some device IDs are hardcoded in usb_devices.c).

 

[Just Checking]

Has anyone reported a decrease in wireless signal strength when going from ver 13E1 to ver 14? I have been doing a lot of testing on some RT-AC66 and RT-AC68 routers and found that when I switched to the ver 14 FW, my signal strength dropped about 10dBi on the 5GHz band and 2-4dBi on the 2.4GHz band. This happened on 2 RT-AC66's and 2 RT-AC68 routers. Yes, I did reboot the routers after I flashed the firmware but I did not do a hard reboot and nvram clear.

I reflashed the ver 13E1 FW and the signal strength went back to where it was originally. I set the power settings at 110mW for the 2.4GHz band and 120mW for the 5GHz band. All the settings remained the same when I changed from one version to the next.

I'm curious since I didn't see anything in the update notes which would indicate that the power output was going to be effected by the FW version change. I went back to the ver 13E1 FW until I get this resolved. It seems to work great on all the routers I have installed it on. It is very stable and reliable - my number one criterion.

 

[john9527]

Has anyone reported a decrease in wireless signal strength when going from ver 13E1 to ver 14?

Nothing changed in the wireless......don't know what to say....

EDIT: Just in case I missed something, I doubled checked against some detailed power measurements I had taken previously...everything was identical (if anything tended toward 1dBm better)

 

[john9527]

Actually quite a fair amount of work would be required to backport the current USB modem support. It's unfortunately not just a matter of upgrading usb-modeswitch (which does help things out), but also the various rom/*.sh and rc/usb*.c (some device IDs are hardcoded in usb_devices.c).

Thanks as always for looking in.....I've been reluctant to change things like this for fear of breaking someone who's working fine today.

 

[Engineer]

Has anyone reported a decrease in wireless signal strength when going from ver 13E1 to ver 14? I have been doing a lot of testing on some RT-AC66 and RT-AC68 routers and found that when I switched to the ver 14 FW, my signal strength dropped about 10dBi on the 5GHz band and 2-4dBi on the 2.4GHz band. This happened on 2 RT-AC66's and 2 RT-AC68 routers. Yes, I did reboot the routers after I flashed the firmware but I did not do a hard reboot and nvram clear.

I reflashed the ver 13E1 FW and the signal strength went back to where it was originally. I set the power settings at 110mW for the 2.4GHz band and 120mW for the 5GHz band. All the settings remained the same when I changed from one version to the next.

I'm curious since I didn't see anything in the update notes which would indicate that the power output was going to be effected by the FW version change. I went back to the ver 13E1 FW until I get this resolved. It seems to work great on all the routers I have installed it on. It is very stable and reliable - my number one criterion.

I flashed and rebooted (no NVRAM reset) and my numbers are identical to what the were before. Good luck.

 

[Just Checking]

Nothing changed in the wireless......don't know what to say....

EDIT: Just in case I missed something, I doubled checked against some detailed power measurements I had taken previously...everything was identical (if anything tended toward 1dBm better)

Thanks for the replies. I noticed this on several routers doing the same thing. I will try flashing the ver 14 again on a couple and see if I can reproduce the results again. It was strange and unexpected.

I noticed that one of the two RT-AC68 routers that I tested had much more signal strength reduction than the other. It was, to me, a new "refurbished" router that I flashed with the ver 14 FW right out of the box. The other RT-AC68 router I tested had every previous version of the Fork FW since ver 2 Beta. Not sure if this is pertinent to the situation, just more information to consider. The RT-AC66's I tested also are ones that are in constant use and have been using the Fork FW since the ver 2 Beta release. I never have had problems of any sort with any of them and they are pretty characterized in terms of expected signal strength at the test location so I am reasonably certain the effect was real. Especially after going back and reflashing the ver 13E1 unto the routers and seeing an increase in measured signal strength.

I will update this post after I flash and retest again. I think I will download the .trx files again to ensure that there was no file corruption which did something strange (if that is even possible).

Since I had this problem, I tried installing the HGGomes 378.55_3 Fork off of RMerlins 378.55_0 FW on the new RT-AC68. I had to flash RMerlins 378.55_0 FW first and do a hard reboot in order to get it to accept the HGGomes Fork. The HGGomez FW behaved as I expected from testing it on a RT-AC87. I'm not really enamored with either RMerlins or HGGomes FW (Just used to John's Fork, since I have it on so many devices now) so I'd like to get this resolved and keep using the latest versions of John's FW.

 

[Just Checking]

The technique I use to assess signal strength is to use a HP Elitebook 8460p Laptop with a dual band WiFi card and I use the "WiFi Finder" app in Mozilla Firefox which provides me with the SSID, mac address, and signal strength in dBi for all the wireless signals received by the antennas in the laptop. I use the same laptop each time since different laptops of the same model can detect different signal strength for the same source. I place the laptop in the same location and orientation each time to ensure that there is no effect from those factors. Since objects moving within the field can affect the measured signal strength, I do most of my testing late at night when most people are gone, or not moving around (They also don't complain over dropped connections and loss of wireless signal when I am flashing FW to the routers and doing reboots).

In this instance, I actually flashed the FW update to the routers by wireless while leaving the laptop in the same position. I measured the wireless signal before and after flashing the new version FW update and rebooting the routers. The signal strength as measured by the "WiFi Finder app" varies over time depending on the surrounding environment and load on the router. Again, that is why I do these updates and tests at night when things are most stable. In this case, the difference in measured signal strength was large enough that I have high confidence that it was statistically significant (certainly passes a T-Test for significance). As an added check, I was evaluating the measured signal strength of several other wireless signals to determine how they changed with time and in relation to the wireless signal from the router I was changing FW on. In effect, I was using the other wireless signals as a control for my measurements and testing. I don't use the MS Windows wireless measurement of link rate because I have never found an actual correlation to signal output of the router.

Granted, I do my testing with little in the way of test equipment specifically designed for this kind of testing. I try to put as much rigor and precision into my testing as I can to compensate for lack of expensive test fixtures. If you have suggestions on how to improve my test methodology, without the use of expensive test equipment, I would be happy to consider them and incorporate them into my test procedures if they are useful.

 

[Deleted member 27741]

Sounds like you have it covered. I have used inssider and it seems ok, I think the new versions are not free however. I actually saw after my post your description of dbi and realized my post was useless and deleted it.

 

[john9527]

@Just Checking - One of the things that struck me is you mentioned a 10dBm change....I think that's over an 8x reduction in power...very significant. Just thinking out loud.....could your Elitebook been on battery vs plugged in when comparing measurements?

EDIT: Another thought just to verify....you are using fixed channels for the measurements.

 

[Ren34]

I just upgrading to V14 from your V13 and now my SSH server isn't working correctly. I am getting this message from my ssh client.
disconnected the session with SshDisconnect.ConnectionLost
Description: FlowSshTransport: received EOF. I already downgraded back to V13 and all is well again. Any ideas?

 

[john9527]

I just upgrading to V14 from your V13 and now my SSH server isn't working correctly. I am getting this message from my ssh client.
disconnected the session with SshDisconnect.ConnectionLost
Description: FlowSshTransport: received EOF. I already downgraded back to V13 and all is well again. Any ideas?

Yes, most likely related to the problem previously reported running scripts remotely with plink. Give it a try again when I post V14E1.

 

[Ren34]

Yes, most likely related to the problem previously reported running scripts remotely with plink. Give it a try again when I post V14E1.

Sounds good. Thanks for the quick response.