Asus RT-AC87R/U - FACTORY F/W Bug Thread

[Poseidon]

Went back to .2678 as the latest fw completely ruined 5GHz wifi performance. It's been a month and still no new fw from Asus. What happened to their once excellent product support?

 

[Dijkstra88]

Went back to .2678 as the latest fw completely ruined 5GHz wifi performance. It's been a month and still no new fw from Asus. What happened to their once excellent product support?

There will be a new beta driver within a week.

 

[chudak]

Ok, bug found. When I choose reboot from the top of the router page it trys to reboot and the internet status light turns red on the router. When I connect to it I cant get into the gui or internet. Only way to fix is to reset by unplugging or the on/off switch.

I see Nightowl has this issues. Can someone else try just hitting the reboot button with the newest firmware. Thanks.

I have the same problem. Have you found a solution?

 

[chudak]

does the same thing when I hit the reboot button, I have to pull power to get it back

Same here, have you found remedy?

 

[lilstone87]

Well just ran into same issue on both the .3561 beta, and the intel beta someone posted recently supposed to help with certain wireless cards. Whenever I turn QoS on, and turn on Traditional QoS. It causes my router to just repeatedly reboot, to the point all I can do is hold the wps button, and do the 30/30/30 reset. I watched the lights on router, and it would never fully boot back up after enabling Traditional QoS. So there is some kind of bug at least in these two firmwares, when enabling the Traditional part of QoS.

 

[Allograft]

there appears to be a bug wherein the client list appears to mistake some wireless clients as having a wired connection. Weird, using 3.0.0.4.376_2769

 

[renderking]

Asus RT-AC87U
Firmware: 3.0.0.4.376_2769

I am also getting "kernel: br0: received packet on vlan1 with own address as source address" in the system log.

Is this a bug or have I done something wrong in the configuration? I upgraded from RT-AC66 and that worked fine. I have reviewed the configuration on both my RT-AC66 and AC87 RT and everything looks fine.

When I get these messages I lose internet connection for 10-20 seconds.
I get these messages whether my switch is connected to port 1 or not but I get them often if my switch is connected and I Watch Netflix via Apple TV approximately every hour.

If I disconnect the switch in port 1 and only use wifi connection I only get the message every 4-6 hours. But there will also be much more activity when the switch is set to the port 1 because in my switch there is Sonos, Philips Hue, Apple TV and PS4 connected only by cable.

Oct 6 06:35:59 kernel: net_ratelimit: 367 callbacks suppressed
Oct 6 06:35:59 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:00 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:01 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:03 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:05 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:07 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:09 kernel: br0: received packet on vlan1 with own address as source address
Oct 6 06:36:11 kernel: br0: received packet on vlan1 with own address as source address

Any updates to this issue? Did disabling off NAT Acceleration resolve? I'm running an AC87 directly behind my modem (Arris TM722G) with a second AC87 in AP mode, both running ASUS's latest stable 3.0.0.4.376_2769 firmware. I had thousands of these error messages in my primary router's syslog when I went to investigate the slowness on the network. One of the cores was running at about 90% and the device was unusually warm. Rebooting primary router solved, but I have noticed about 4 additional "kernel: br0: received packet on vlan1 with own address as source address" in the course of 24 hours, so the ghost still lives.

Equipment is setup at a remote site, so I don't have easy access to physically trouble shoot, but I do have remote access. Any additional insight would be appreciated. Logs attached in case there's any value.

 

[Haggster]

Adaptive QoS Bandwidth Monitor - "empty" option missing now

Hi,

I realized that in the QoS Bandwidth Monitor the "empty" tag is missing. This is the case with the current official September 2014 firmware and with the December 26 2014 unofficial firmware.

This means that once you tag one of the connected devices as either "Highest", "High", "Medium", "Low" or "Lowest", you cannot untag them anymore. Nothing fixes this.

I tried:
- using another browser (Firefox instead of Chrome)
- rebooting
- flashing the latest beta firmware
- shutting QoS off and turning it back on

Even the official documentation mentions the option to use the "empty" tag to bring a tagged client back to neutral QoS state.

Below are two pictures showing how it is supposed to look (taken from a web UI based on an apparently older firmware) and a screenshot taken from my router with the current Dec. 26 beta firmware (looks the same with the current official firmware):

Can anyone confirm this?


EDIT: just for reference, a link to the official documentation which states "Drag and drop “empty tag” to device to cancel device’s priority setting.". Oddly enough, their screenshots are missing the Empty tag as well and the document can only be found in google's cache while this FAQ entry seems to have been otherwise removed:

http://webcache.googleusercontent.com/search?q=cache:vaMEXnPewBkJ:www.asus.com/support/FAQ/1008717/

 

[scavenger]

Can anyone confirm this?

I can confirm it is no longer there in 3754, and I also recall it being there a few firmwares back. I used it more than once to clear priorities.

However I'm not sure how the Adaptive QoS treats an unclassified client. Maybe the empty tag was redundant, and that it was functionally identical to just declaring a client as "Medium" priority?

I would think you can't really have an uncategorized client operate outside of the QoS engine, or it would defeat the purpose, so I was never sure what state an 'empty' host was in.

 

[arachnida]

Quite honestly, I don't understand why so many people are having such issues, while a lot of us don't have any. Personally, the RT-AC87U has been my main router since before its release, and all my wireless devices are working just fine with it so far:

• One laptop with an Intel AC7260 (5 GHz)
• One Nexus 7 (2.4 GHz)
• One Nexus 4 (2.4 GHz)
• One Asus Transformer T100 (some Broadcom chip, 5 GHz)
  • 
    
    The fact that different people have different experiences lead me to believe that any issue might be client or environment specific.

• • 
    
    I will try out once more without connecting any iDevices, NEST or SonoS, only Nexus 7 2012.
    
    But reading the two different forums most people running the AC87 in AP mode seem to have the packet storm issue. Do you also have an AC87 in AP mode or only in "normal" mode? Someone having the very same problems also has SonoS in his network. Coincidence? (The latest updates of SonoS have changed their mesh network so that 2.4 GHz Wifi can also be used.)

 

[Haggster]

I can confirm it is no longer there in 3754, and I also recall it being there a few firmwares back. I used it more than once to clear priorities.

However I'm not sure how the Adaptive QoS treats an unclassified client. Maybe the empty tag was redundant, and that it was functionally identical to just declaring a client as "Medium" priority?

I would think you can't really have an uncategorized client operate outside of the QoS engine, or it would defeat the purpose, so I was never sure what state an 'empty' host was in.

Thanks a lot for your reply. As far as I understand, all clients that are classified get *all* of their traffic prioritized (if set to higher priority) or throttled (if set to lower). Unclassified clients get treated depending on the QoS rules for the different kinds of traffic, i.e. if you don't want to prefer one single client over others, you can just leave all clients unclassified and let the QoS do its work according to its settings. Setting a client to medium will always deprive it of some of its bandwith as far as I understand. So for a clean implementation of the QoS rules it should be possible to remove any classification via the "empty" tag.

Does anyone else have any thoughts on this or know where the button/tag went?

 

[ogre]

Updated to 3754 without problems, but when I reset the router to factory settings with long press on reset, the following bug appears:

The system settings page where I',m supposed to set the timezone says

* Remind: The System time zone is different from your locale setting.

But it's impossible to set the timezone, since the checkbox for daylight savings (that is there on my RT-AC66u) is not in the GUI. This also means that I can't change the admin name or password, since it is on the same page.

Anyone else seen this?

 

[RMerlin]

Updated to 3754 without problems, but when I reset the router to factory settings with long press on reset, the following bug appears:

The system settings page where I',m supposed to set the timezone says



But it's impossible to set the timezone, since the checkbox for daylight savings (that is there on my RT-AC66u) is not in the GUI. This also means that I can't change the admin name or password, since it is on the same page.

Anyone else seen this?

The new firmware seems to have a broken System page. There's a Javascript error that prevents the page from fully loading, related to their new timezone handling code.

 

[ogre]

New FW ending with 3885 released - still broken system page