Asuswrt-Merlin 374.40 Beta 2 is available

[wifi_reef]

I'm currently on 40.3b testing with RT-AC68U. I'm getting reboot issue as described in the thread. I've looked in to the log - nothing there. However I've noticed when this had happened I had tries to stream pictures over media bridge to TV Panasonic.

Environment: RT-AC68U / WL-500W (Media Bridge) / Panasonic TV wired to WL-500W / Some more units wired to RT-AC68U / up to five 2.4 GHz clients / up to four 5.0 GHz clients.

Now I can repeat any time router's reboot with this procedure:

1. WiFi on Sony Xperia V - Android 4.3 from Sony,
2. connection established and operating (the same with 2.4 / 5.0),
3. Open Album,
4. Choose picture,
5. Use Throw function (this allow to stream multimedia) and pointing at Panasonic TV,
6. streaming starts,
7. ROUTER reboots !!

Sometimes it reboots with first picture sometimes at the next one - however it will reboot every time when followed procedure.

This is not happening in Meriln's 38.2 fw. Throw works flawlessly with 38.2 fw. I hope it can help with solving RT-AC68U rebooting issue after 583 GPL.

 

[linda.oh]

Hi, I'm using ac66u Firmware Version3.0.0.4.374.38_2

May I know what this mean?

Mar 10 13:55:27 rc_service: httpd 305:notify_rc restart_wlcscan\

Thanks

 

[RMerlin]

Hi, I'm using ac66u Firmware Version3.0.0.4.374.38_2

May I know what this mean?

Mar 10 13:55:27 rc_service: httpd 305:notify_rc restart_wlcscan\

Something that you did on the web interface told the router to do a scan to see which wireless networks were available. Can happen amongst other things if you use the Site Survey page, or are trying to configure a repeater. It's just standard logging, nothing unusual there.

 

[Just Checking]

Problem with Temp Monitor & Site Survey

Router: Asus RT-AC66R with 9db High Gain Antennas
Firmware: 374.40 Beta 2
Problem #1:
Temperature Monitor will not display Temperature in Fahrenheit after applying defined field. The default is in C. and I can live with that but F. is more sensitive in the readout. I have tried to set the default to F. but after logging off from the GUI and logging back into the router, the display is back to C. I have to change it each time I want to see the information in F.. I am also interested in getting a log set up of the temperature. I see no way to do this.

This is a very minor problem and does not really detract from having this temp monitoring feature which I really like.

Problem #2:
Site Survey is another excellent feature that I really like. With the 374.40 Beta 2 version, when doing a scan in Passive mode, the indicator of signal strength shows almost all the radios in the area at 5 bar signal strength. If I rescan in Active mode, the number of surrounding radios goes down significantly (understandable) and, the reported signal strength also goes down for the radios that are detected. If I log off the router GUI and log back in then go back to Site Survey and do a rescan in Passive Mode, I often pick up other radios but the ones I pick up from the first scan often go back to more "normal" radio output (0-3 bars). I seem to remember not having this problem on the 374.39 firmware version but I could be wrong about that.

I live in a multi-family dwelling with three other WiFi radios (AP's) from other tenants. Those I pick up at full signal strength at all times in Active and Passive search modes. The building is sheathed in metal and there is a lot of metal in the outer building walls. This does an effective job of blocking signals in and out of the building. There should be no way for me to receive 5 bar signal strength (2.4GHz band) from buildings located 50 meters away. My routers radio regularly detects signals from buildings 100+ meters away.

Is there a way to improve this firmware to give an actual reading of signal strength (like the WiFi add-on for Mozilla Firefox)? More importantly, is there a way to improve the, to me, unreliable nature of the survey scan? I regularly detect 17+ different radios on the 2.4GHz band so it is important for me to understand the radio signals and the band which they reside on (Love this feature) to optimize my routers performance. Fortunately, only my next door neighbor outside the building has a router currently using the 5GHz band and I can easily choose channels for each of my routers to not get interference from his signal. It is very difficult to do that on the 2.4GHz band.

 

[L&LD]

Any reason you're not on the 374.40 Final?

 

[Just Checking]

374.41 alpha

Is the 374.41 alpha the latest version? I saw reference to that but have not found it to download yet and try. I was looking through the revision notes and saw that it had different base algorithms to improve speed. Is that true? I have not done any test between the 374.39 and the 374.40 fw versions but the 374.40 versions seem slower than the 374.39 version.

My fw version may be 374.40 beta3. It was the latest as of 08 March 2014 when I downloaded and installed it.

I have only switched to the Asus Merlin fw with 374.39. I used stock fw before that. I only switched because of need to get features like VPN but I found a lot more features that are desirable. I would only switch back if I get improvements in wireless speed from the stock firmware. So far, I am not seeing any problems with lost connections or hangups with the Asus Merlin fw that indicate unreliability. Just a few minor foibles that I am hoping can be worked out.

 

[Just Checking]

FW ver 374.40 released - Same Problem w. Temp

I downloaded the released version of 374.40 and the Temp still defaults to C. when I log off and log back in.

It isn't a big deal. It is just a little irritation.

 

[Just Checking]

Site Survey FW for 374.40 Released works Better

I checked out the Site Survey feature on the 374.40 released FW and it seems to work better than the Beta version that I tried. Now there isn't the large number of radios all having 4 or more bars on the passive scan.

I noticed also that when I hover the cursor over the WiFi signal strength indicator I get a pop up of the % signal. I'm not sure what that equates to in terms of signal strength or db. Does anyone have a clue about that?

I like the feature of being able to left clic on the ssid name and get the mac address and the wireless radio manufacturer. I noticed that some of the radios have the radio manufacturer disabled or not present. I think this is a excellent security feature. Does anyone know how to disable this on a router. It isn't because the router is set to a silent ssid. I get the mac address and router mfg. on radios that are not broadcasting a ssid. I see nothing from some radios that broadcast a ssid and a mac address. Could this be a feature built into some mfg radios (like they build in broadcasting the radio mfg and address along with the mac address)? I am just curious how this is done and if it could be built into a next gen FW build to really stealth these Asus routers.

 

[RMerlin]

I noticed also that when I hover the cursor over the WiFi signal strength indicator I get a pop up of the % signal. I'm not sure what that equates to in terms of signal strength or db. Does anyone have a clue about that?

Here's how the firmware calculates it:

if (info->RSSI >= -50)
     apinfos[idx_same].RSSI_Quality = 100;
else if (info->RSSI >= -80)     // between -50 ~ -80dbm
     apinfos[idx_same].RSSI_Quality = (int)(24 + ((info->RSSI + 80) * 26)/10);
else if (info->RSSI >= -90)     // between -80 ~ -90dbm
     apinfos[idx_same].RSSI_Quality = (int)(((info->RSSI + 90) * 26)/10);
else                                    // < -84 dbm
     apinfos[idx_same].RSSI_Quality = 0;

I like the feature of being able to left clic on the ssid name and get the mac address and the wireless radio manufacturer. I noticed that some of the radios have the radio manufacturer disabled or not present. I think this is a excellent security feature. Does anyone know how to disable this on a router. It isn't because the router is set to a silent ssid. I get the mac address and router mfg. on radios that are not broadcasting a ssid. I see nothing from some radios that broadcast a ssid and a mac address. Could this be a feature built into some mfg radios (like they build in broadcasting the radio mfg and address along with the mac address)? I am just curious how this is done and if it could be built into a next gen FW build to really stealth these Asus routers.

I've never observed this. My guess would be those radios are broadcasting a MAC that isn't registered in the OUI database, causing the lookup to fail. You would have to forge your wireless MACs to achieve a similar result.

 

[RMerlin]

I downloaded the released version of 374.40 and the Temp still defaults to C. when I log off and log back in.

It isn't a big deal. It is just a little irritation.

If you're referring to the Performance Tuning page, the setting is stored in a cookie. If the setting gets reverted back to C, then it means you are not keeping the cookie.

If you are referring to the Sysinfo page, that page is hardcoded in Celsius.

You won't get better granularity with Fahrenheit BTW, since the temperature is reported in Celsius by the HW. The Fahrenheit value is calculated from the Celsius reading.

 

[Just Checking]

Any way to report actual db

Here's how the firmware calculates it:

if (info->RSSI >= -50)
     apinfos[idx_same].RSSI_Quality = 100;
else if (info->RSSI >= -80)     // between -50 ~ -80dbm
     apinfos[idx_same].RSSI_Quality = (int)(24 + ((info->RSSI + 80) * 26)/10);
else if (info->RSSI >= -90)     // between -80 ~ -90dbm
     apinfos[idx_same].RSSI_Quality = (int)(((info->RSSI + 90) * 26)/10);
else                                    // < -84 dbm
     apinfos[idx_same].RSSI_Quality = 0;

I've never observed this. My guess would be those radios are broadcasting a MAC that isn't registered in the OUI database, causing the lookup to fail. You would have to forge your wireless MACs to achieve a similar result.

If I am interpreting this coding correctly, the fw does get an actual reading of db then algebraically manipulates that to give a % signal. Can the code be changed to just report the actual db? Or, code added to report the actual db as well as %?

This code would indicate that any detected signal with strength greater than -50db is reported as 100% signal. That doesn't seem to be reasonable. The tables of throughput on most routers, especially the RT-N66, RT-AC66, and the RT-AC68 have significant wireless transfer rate fall off when the signal strength falls below -30db. Yes, I know that the code has to be arbitrarily set for the # of bars signal strength but this seems to give people a fall sense that the routers should perform better than they actually will.

Just my US$0.02 on that.


Regarding forging, or altering, the mac address, how would one go about doing that? Would that be a feature which could be put in as an option (Stealth Radio Mode)?

 

[RMerlin]

If I am interpreting this coding correctly, the fw does get an actual reading of db then algebraically manipulates that to give a % signal. Can the code be changed to just report the actual db? Or, code added to report the actual db as well as %?

If I do any change to this section of the code, it will break other functionalities, such as the repeater connection assistant.

Yes, I know that the code has to be arbitrarily set for the # of bars signal strength but this seems to give people a fall sense that the routers should perform better than they actually will.

The site survey is for seeing how strong your neighbour's signals are coming in, to help determining which channel is the least crowded. For this use, a percentage is adequate. Plus, mom & dad users will have no idea how to deal with those negative numbers, while they can easily understand a percentage. So if it was a choice between signal level and percentage, percentage is what makes the most sense for the intended audience.

Regarding forging, or altering, the mac address, how would one go about doing that? Would that be a feature which could be put in as an option (Stealth Radio Mode)?

It might be controllable through an nvram setting, but I never tried, and I don't know either if it will be overwritten by the firmware at boot time.

 

[Just Checking]

If I do any change to this section of the code, it will break other functionalities, such as the repeater connection assistant.


It might be controllable through an nvram setting, but I never tried, and I don't know either if it will be overwritten by the firmware at boot time.

First, Thanks for answering my questions and responding to my comments. I really appreciate your considerate answers. I don't have access to the coding and have not spent the 1000's of hours evaluating the coding to understand the interaction that you have. When I Blithely make a request about changes, they are because they are features that I would like to see but, also to help me understand what is going on. Your answers are helpful in that way even if you say you cannot do something.

The site survey is for seeing how strong your neighbour's signals are coming in, to help determining which channel is the least crowded. For this use, a percentage is adequate. Plus, mom & dad users will have no idea how to deal with those negative numbers, while they can easily understand a percentage. So if it was a choice between signal level and percentage, percentage is what makes the most sense for the intended audience.

I somewhat disagree that mom & dad users are going to go deeply into this firmware. I contend that only people who are pretty serious about their networking are going to load your open source firmware in the first place (I only recently did this and I have a more sophisticated network and network knowledge than anyone I know personally who doesn't do this every day for a living.). I also consider myself to be a neophyte at this. This site survey tool is powerful and if a few of the functions could be tweaked, it would be really powerful.

My real concern about this is the fact that any signal stronger than -50dbm is reported as 100% strength. If this is true, than it makes this portion of the tool somewhat worthless for many purposes including the channel selection. Those great charts that Mr. Higgins spends a lot of time generating indicate that data transfer rates decrease precipitously when signal strength falls below -20dbm and certainly below -30dbm. Would it not make much more sense to set the 100% signal strength limit at -30dbm or, even better, at -20dbm? That would only require a change in the value of the coding. I am unfamiliar with any of the coding to understand whether this would disable, or inhibit, the repeater connection assistant.

This is just my wishful thinking since the coding can't be altered without significant other coding changes.


Again, I appreciate your answer on the mac address changes. It was another wishful thought on how to further stealth my network and make it more resistant to intrusions. People who drive around and try to hack into wireless connections have a easier time when they know the equipment they are trying to enter. Knowing the wireless radio mfg. tells them the router and a starting IP address to enter the router. It also makes it easier to enter because most people do not change the administrator name even if they do change the password (I know quite a few people who leave the administrative password blank on their routers because they do not want to be bothered with remembering the password). Also, some router mfg. (Buffalo, for one) limit the character length and character set on their passwords. It makes it easier for a hacker who is knowledgeable to get the administrative name and password if they know the router manufacturer.

Finally, if the router mac address could be spoofed, or altered on ISP's that have a dynamic IP address, it would be a great tool to help prevent tracking of your usage by companies like Google since both the network IP address and network mac address can vary every time the router is reset (or whenever the router fw is set to reset). Just my paranoia showing through here. I think it would be a very cool feature. I think it would be extremely difficult to implement with VPN but great for people just wanting to stealth their general networks. Just my US$0.02 on that too.

 

[RocketJSquirrel]

I somewhat disagree that mom & dad users are going to go deeply into this firmware. I contend that only people who are pretty serious about their networking are going to load your open source firmware in the first place (I only recently did this and I have a more sophisticated network and network knowledge than anyone I know personally who doesn't do this every day for a living.). I also consider myself to be a neophyte at this. This site survey tool is powerful and if a few of the functions could be tweaked, it would be really powerful.

I can vouch that mom & pop don't go into the firmware. My neighbor directly across the street (with his router as close as he could possibly place it opposite mine) got a new RT-AC66U which was apparently installed by his adult son. Son had left it on Auto channel on both radios, so it was prancing all up and down the spectrum interfering with everyone nearby.

Luckily, he's a friendly neighbor and allowed me to tweak his router's settings to stay put on a single channel per band. He watched me do it and had absolutely no idea what I was doing or how I even got into the router's GUI.

Of course, he was running stock Asus firmware. So when I went to tweak his router, I carried my laptop running inSSIDer. So I was able to choose a channel for him based on his location. If you're interested in seeing the neighboring signals, inSSIDer was made for that and gives you the RSSI in negative dBs just like you want. So that's what I'd recommend for that purpose.

 

[Just Checking]

Thanks

Thanks. I will check inSSIDer out. I like having the feature on the router itself which is configured with the radios and antennas that would have potential interference. The antennas and wireless adapter cards on my laptops are neither as sensitive or powerful.

Mozilla Firefox has an add-on called WiFi Finder which gives the mac addresses and signal strength. I have it on all of my laptops which have wireless adapter cards. Unfortunately, it doesn't tell the channel which the radios are on. if inSSIDer does that, it would definitely improve functionality.

 

[Just Checking]

374.40 Final Decreased Wireless Transfer Rate?

I was doing some transferring of about 120GB of media files yesterday and noticed that the transfer rate maxed out at about 21MB/sec on the 5GHz band using a Asus RT-AC66R router and Asus-Merlin Firmware ver 374.40 final. When I was using stock Asus firmware (version loaded in the router as coming from the mfg.) I was routinely getting 25MB/sec - 28MB/sec.

The router and the wireless client are in the same position for each transfer. The wireless environment have not really changed either (according to the site survey tool).

Previous posts have noted that the 374.39 and 374.40 versions seem to have slowed down wireless transfer speeds. One poster insisted that 374.35 was much faster. Has anyone else noticed this?

I was looking at the notes on 374.41 alpha builds which, to me, indicate that a different kernel is being evaluated which might improve speed. Merlin, can you comment on this?

Should I go back to the ver. 374.35 firmware and wait for the 374.41 to come out?

 

[RMerlin]

I was doing some transferring of about 120GB of media files yesterday and noticed that the transfer rate maxed out at about 21MB/sec on the 5GHz band using a Asus RT-AC66R router and Asus-Merlin Firmware ver 374.40 final. When I was using stock Asus firmware (version loaded in the router as coming from the mfg.) I was routinely getting 25MB/sec - 28MB/sec.

I'm using the wireless driver from Asus's 374.2xxx or 374.4422 release (I forgot which). The latest driver was causing quite a few users to significantly lose in range/performance.

So for all intent ant purposes, you should get the same performance as Asus's stock firmware on the same driver.

Previous posts have noted that the 374.39 and 374.40 versions seem to have slowed down wireless transfer speeds. One poster insisted that 374.35 was much faster. Has anyone else noticed this?

Always check which router model people are talking about (this is why it's critical for people to mention this). Those people are mostly referring to the RT-N66U.

I was looking at the notes on 374.41 alpha builds which, to me, indicate that a different kernel is being evaluated which might improve speed. Merlin, can you comment on this?

No change in wifi driver in 374.41, since there has been no new code from Asus merged in it. The only real change was an updated version of miniupnpd.

Wifi driver is 100% function of what Asus releases. It's closed source, I can do zero changes to it.

 

[RMerlin]

I somewhat disagree that mom & dad users are going to go deeply into this firmware. I contend that only people who are pretty serious about their networking are going to load your open source firmware in the first place (I only recently did this and I have a more sophisticated network and network knowledge than anyone I know personally who doesn't do this every day for a living.). I also consider myself to be a neophyte at this. This site survey tool is powerful and if a few of the functions could be tweaked, it would be really powerful.

I still target less experienced users than Tomato or DD-WRT. A number of users only flashed this firmware for the bugfixes, or for some very user-friendly features such as traffic monitoring (one of my own customers only wanted to be able to track his children's Internet usage, as he was getting monthly overage fees from his ISP due to excessive traffic).

My real concern about this is the fact that any signal stronger than -50dbm is reported as 100% strength. If this is true, than it makes this portion of the tool somewhat worthless for many purposes including the channel selection. Those great charts that Mr. Higgins spends a lot of time generating indicate that data transfer rates decrease precipitously when signal strength falls below -20dbm and certainly below -30dbm. Would it not make much more sense to set the 100% signal strength limit at -30dbm or, even better, at -20dbm?

When it comes to interference, if you have anything that is stronger than -50 dBm, then you should be ensuring you aren't overlapping with that AP's channel. For this use, this information is accurate enough. This has nothing to do with the throughput degradation related to signal strength reported in Tim's testing - he's talking about client-to-AP performance, while I'm talking about interference impact. For the aspect that would match Tim's discussions, the information is available in great details on the Wireless Log page, which will report signal strength of your clients.

The goal of the Site Survey is not to replace more specialized tools such as InSSIDer, which is typically used just as much for evaluating your own network as checking on your neighbour's networks. It's just a quick interface to determine how likely your neighbours are to interfere with your own network.

 

[teh_g]

How has stability been for people with the AC-68U? I saw some initial stability issues, but wanted to see if this would impact me too much...

 

[Maldades]

How has stability been for people with the AC-68U? I saw some initial stability issues, but wanted to see if this would impact me too much...

5 days without any reboot.
I have installed the 40 Alfa.