[Release 384/NG] Asuswrt-Merlin 384.5 is now available

[RMerlin]

I run 384.3-beta3 on my AC3200 as it’s the only new release that 2.4ghz is usable for me. I have tried the new stock 382_50010 for a week or so and can confirm no drops with that. It’s definately an Asus issue not Merlins firmware so I’m hopeful that the GPL for 50010 will be released at some point soon and Merlin is kind enough to release a new version with this included

382_5010 is not the latest, 382_50470 is.

384.5 uses the wireless driver from 382_50010.

 

[Makaveli]

Cooler are not that good solution, additional power needed, dust gets inside, doesnt help the processor with no direct contact to the cooler plate to get the heat off.

Have to disagree with you on this.

I got a 17-19c cpu temp drop with a laptop cooling pad.

A can of compress air easily cleans it.

 

[jernau]

382_5010 is not the latest, 382_50470 is.

384.5 uses the wireless driver from 382_50010.

Apologies I did mean 50740 which was released in May sorry for the confusion

 

[Grisu]

Have to disagree with you on this.

I got a 17-19c cpu temp drop with a laptop cooling pad.

A can of compress air easily cleans it.

Disagree you too!

This device is disigned and planed to have >70°C temps, so what do you get with additional coolers?
Just nothing but more power consumption, maybe noise and dust and cleaning every year.

Its a great difference to have it running passiv cooling at 75°C which is pretty good forever or runnnig at >85°C nearby its design limits.
Maybe even I am too pedantic ...
Just want to say if it is on some devices possible without additional coolers to run it at 75° of course is better than to have >85°C regardless real life benefit (possibility of faults maybe reduced).

You can even cool it to <0°C, but just for nothing than additional power consumption and in your poor opinion good 'values' which aren't good design values for technicians.

 

[iManuB]

I still have, sometimes, this HTTPD crash ...

What can be caused by? Hardware?

May 23 12:39:45 kernel: httpd[797]: unhandled level 3 translation fault (11) at 0x005c322c, esr 0x92000007
May 23 12:39:45 kernel: pgd = ffffffc0156a9000
May 23 12:39:45 kernel: [005c322c] *pgd=00000000156a8003, *pud=00000000156a8003, *pmd=00000000153b4003, *pte=0000000000000000
May 23 12:39:45 kernel: CPU: 1 PID: 797 Comm: httpd Tainted: P           O    4.1.27 #2
May 23 12:39:45 kernel: Hardware name: Broadcom-v8A (DT)
May 23 12:39:45 kernel: task: ffffffc01934aa80 ti: ffffffc01500c000 task.ti: ffffffc01500c000
May 23 12:39:45 kernel: PC is at 0xf6c59aa0
May 23 12:39:45 kernel: LR is at 0x25f28
May 23 12:39:45 kernel: pc : [<00000000f6c59aa0>] lr : [<0000000000025f28>] pstate: 20000010
May 23 12:39:45 kernel: sp : 00000000ffd263b8
May 23 12:39:45 kernel: x12: 00000000000876e4
May 23 12:39:45 kernel: x11: 00000000000746b5 x10: 0000000000000001
May 23 12:39:45 kernel: x9 : 00000000005a4b30 x8 : 0000000000000000
May 23 12:39:45 kernel: x7 : 00000000005b4ac0 x6 : 0000000000597280
May 23 12:39:45 kernel: x5 : 00000000005c3220 x4 : 00000000005c3220
May 23 12:39:45 kernel: x3 : 00000000ffffffff x2 : 00000000f6b057a8
May 23 12:39:45 kernel: x1 : 0000000000000000 x0 : 00000000005c3220
May 23 12:40:11 watchdog: restart httpd
May 23 12:40:11 rc_service: watchdog 805:notify_rc stop_httpd
May 23 12:40:11 rc_service: watchdog 805:notify_rc start_httpd
May 23 12:40:11 RT-AC86U: start httpd:80

Has anyone encountered this error? What is it due to?

 

[Grisu]

Has anyone encountered this error? What is it due to?

Have it on stock asuswrt, on merlin with 68U and 86U even with new 68U beta from today ...

So general problem nothing merlin can or should solve.

 

[agilani]

So now that the AC88 has been stable and not requiring any reboots, it looks like the webui cpu usage may be reporting wrong? I show core2 at 100%, but when i look at the processor usage, nothing is higher than 0.8%

Mem: 207496K used, 307688K free, 2092K shrd, 3616K buff, 14300K cached
CPU: 0.6% usr 1.3% sys 0.0% nic 46.4% idle 46.7% io 0.0% irq 4.7% sirq
Load average: 5.21 5.08 5.06 1/120 16731
PID PPID USER STAT VSZ %VSZ CPU %CPU COMMAND
329 1 admin S 5536 1.0 0 0.8 /sbin/netool
399 1 admin S 5880 1.1 1 0.5 httpd -i br0
328 1 admin S 2084 0.4 0 0.3 protect_srv
906 1 admin S 908 0.1 0 0.3 miniupnpd -f /etc/upnp/config
352 1 admin S 1364 0.2 1 0.1 /bin/eapd
432 1 admin S 6264 1.2 1 0.0 mastiff
446 432 admin S 6264 1.2 1 0.0 mastiff
447 446 admin S 6264 1.2 1 0.0 mastiff
429 1 admin D 5936 1.1 1 0.0 networkmap --bootwait
403 1 admin S 5640 1.0 1 0.0 watchdog
336 329 admin S 5536 1.0 1 0.0 /sbin/netool
337 336 admin S 5536 1.0 0 0.0 /sbin/netool
1 0 admin S 5528 1.0 1 0.0 /sbin/preinit
320 1 admin S 5520 1.0 0 0.0 /sbin/wanduck
858 1 admin S 5516 1.0 1 0.0 bwdpi_wred_alive
.
.
.
.

 

[Mark112887]

I'm back with an update on my AC 3200. After a little more than 48 hours of completely solid uptime, my 2.4 GHz network has just recently started dropping connection every 30 seconds to one minute. After about 30 seconds my printer seems to reconnect then go out again a few seconds later. While all of this is going on my 5 GHz networks are not affected at all.

Here is my log file,

 

[agilani]

I'm back with an update on my AC 3200. After a little more than 48 hours of completely solid uptime, my 2.4 GHz network has just recently started dropping connection every 30 seconds to one minute. After about 30 seconds my printer seems to reconnect then go out again a few seconds later. While all of this is going on my 5 GHz networks are not affected at all.

Here is my log file,

This is almost identical to what i saw. Some sort of DHCP loop over and over. Try turning off your printer or move it to the 5ghz band if it supports it and see if that stabilizes your 2.4Ghz. I tracked down every device that showed these symptoms and moved them off of my network to stabilize 2.4ghz band. I actually had 4 or 5 devices showing these symptoms. You only seem to have the one. Another option to try is to statically assign an address in the printer to see if that helps and exclude it from your dhcp address range.

May 26 00:46:42 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:47:02 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:47:02 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:47:22 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:47:22 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:47:42 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:47:42 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:48:59 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:48:59 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:49:19 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:49:19 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:49:44 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:49:44 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:50:20 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:50:20 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:50:51 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:50:51 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:51:06 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48
May 26 00:51:06 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer

 

[Jon555]

I'm back with an update on my AC 3200. After a little more than 48 hours of completely solid uptime, my 2.4 GHz network has just recently started dropping connection every 30 seconds to one minute. After about 30 seconds my printer seems to reconnect then go out again a few seconds later. While all of this is going on my 5 GHz networks are not affected at all.

Here is my log file,

Sounds similar to the issues I was having on my AC3100 a couple versions of 384 ago (haven’t tried the new one yet). 2.4 network would work great for ~2-6 hours after a reboot, and then immediately begin dropping every 30 seconds. One may suggest to move devices to the 5 GHz network, but this was a problem I had never encountered before when using the legacy 380 firmwares. Apologies if it sounds as though I’m complaining; I’m not (it’s just my experience so far with 384). I’m more than confident (and greatful) that RMerlin will sort any issues like this out (assuming he even can and it’s not an ASUS issue).

 

[joegreat]

Some sort of DHCP loop over and over.

May 26 00:46:42 dnsmasq-dhcp[302]: DHCPACK(br0) 192.168.1.4 6c:3b:e5:00:0f:48 Printer
May 26 00:47:02 dnsmasq-dhcp[302]: DHCPREQUEST(br0) 192.168.1.4 6c:3b:e5:00:0f:48

This is a know issue for long time and need to be solved by the device vendor - look for a firmware/software update for your device. In the past Apple devices where known for this (and Apple refused to fix it).

The simplest way to get rid of this useless messages in the log is to disable them under LAN / DHCP Server - Option "Hide DHCP/RA queries" = Yes.

 

[Paulo Pereira]

Hello, is there a way to add "Set AP isoleted" to the Guest Network??? My router RT-AC5300 is working on Access Point (AP) Mode. It was interesting this features, because who access to normal wifi should be able to connect to other machine in network, but who access to the Guest Network is temporary user and don't be able to access to the others machine's in the network. Sorry for the english and Thank's in advance.

 

[eddiez]

Thats a way too hot, yes it will work and probably wont overheat for some years, but its not good!

My 68U was 85°C as well and after improoving thermal contact between processor and thermal plate it is now down to 75°C, dispite its running 1200/800MHz now and before only 800/666 and that makes about 3-5° more temp.
My 86U got under 75° out of the box, so no need to open it.

Cooler are not that good solution, additional power needed, dust gets inside, doesnt help the processor with no direct contact to the cooler plate to get the heat off.
In my 68U I had to bend the plate a little bit in the range where the processor is sitting. It is pressed down on the sides and bending up, so where needed space between processer to plate.
Somewhere I found a nice video of it where the problem is shown, but did it my way years before ...

Please check the designed operating temps for this SOC. It will show you that these temps (80/85) are perfectly within critical boundaries...

This discussion can also be found way back in the forum. If lower temps make you feel more comfortable, just improve the airflow to the router with a $5 sub fan. But effectively these soc's only become critical with change of degraded lifespan when running 100+ degrees continuously.

Verstuurd vanaf mijn SM-G965F met Tapatalk

 

[Grisu]

Please check the designed operating temps for this SOC. It will show you that these temps (80/85) are perfectly within critical boundaries...

This discussion can also be found way back in the forum. If lower temps make you feel more comfortable, just improve the airflow to the router with a $5 sub fan. But effectively these soc's only become critical with change of degraded lifespan when running 100+ degrees continuously.

Verstuurd vanaf mijn SM-G965F met Tapatalk

you dont look further in summer when room temperature will rise from 20 to over 30°C and maybe heavy load too, router temps will rise linar up and than over 90° it is hitting design limits.
Nothing is good operated at its rating limits all time!
Then on higher temps semiconductors will have less resistance what means more current as voltage is regulated and unchanged and so power consumption will increase even more.
All power is transformed to heat which must have to be cooled anyhow.

Asus did design it well within limits, but many devices are really badly put together with bent circuit board, so leaving space between processor and plate.
I have seen it on two of my previous 68U !

All I want to say is NOT to add a cooler fan but much better to resolve this production problem by repairing and correcting it.
They can easily be opened with a credit card after remooving screw on back.

It is one thing to change the design by adding hardware and another to correct faulty assembling.

 

[eddiez]

you dont look further in summer when room temperature will rise from 20 to over 30°C and maybe heavy load too, router temps will rise linar up and than over 90° it is hitting design limits.
Nothing is good operated at its rating limits all time!
Then on higher temps semiconductors will have less resistance what means more current as voltage is regulated and unchanged and so power consumption will increase even more.
All power is transformed to heat which must have to be cooled anyhow.

Asus did design it well within limits, but many devices are really badly put together with bent circuit board, so leaving space between processor and plate.
I have seen it on two of my previous 68U !

All I want to say is NOT to add a cooler fan but much better to resolve this production problem by repairing and correcting it.
They can easily be opened with a credit card after remooving screw on back.

If you would have read correctly, the limits are very far away from the max thermal SOC design which is 125 degrees (you apparently did not check yourself )

My router runs currently (and has been for 4 years) in a closed space (limited airflow), with an ambient temperature of 28/29 degrees. The router temp varies between 84 (no load) and 87 degrees (heavy load). I don't know what kind if load you wsnt to put in the router but there hasn't been any report of a router even reaching 100 degrees. So please stick to the facts.

Verstuurd vanaf mijn SM-G965F met Tapatalk

 

[zoggy]

quick update:

brand new rt-ac88u a few weeks old, while on on 384.x I kept having 2.4ghz issues where that radio would drop out and anything that was using it would try to reconnect and fail until you turn the radio off then on / reboot the device. 5ghz would be fine and the led for 2.4ghz was still on. this problem is prevalent across multiple devices through these forums.. some people swear its hw issue some people say its a bug in the firmware. I gave 384.5 a chance and within a few hours the issue happened. I decided to try downgrading to 380.70 and I can report now that everything has been stable for 4.5 days now. going to monitor for another week to confirm it really is 'fixed' by downgrading software. for completeness: 33 devices connected, 15 are on 2.4ghz spectrum.

 

[bill1228]

If you would have read correctly, the limits are very far away from the max thermal SOC design which is 125 degrees (you apparently did not check yourself )

My router runs currently (and has been for 4 years) in a closed space (limited airflow), with an ambient temperature of 28/29 degrees. The router temp varies between 84 (no load) and 87 degrees (heavy load). I don't know what kind if load you wsnt to put in the router but there hasn't been any report of a router even reaching 100 degrees. So please stick to the facts.

Verstuurd vanaf mijn SM-G965F met Tapatalk

I tend not to worry too much about temps. Folks worry about shortening the life of the router. I figure if it fails after 2 or 3 years because of temperature, there is going to be a newer, better, faster router that I am wanting an excuse to buy anyway. CPU temp on my rt86u is about 73 degrees right now. If it got up to 85 or 90 I would not worry. I remember when running a Linksys WRT1900AC V1 that Linksys had the fan on temp as 85 and off at 80. Not the most wonderful router I might add.

 

[kfp]

Load average: 5.21 5.08 5.06

That looks pretty high.

 

[kfp]

This is almost identical to what i saw. Some sort of DHCP loop over and over.

I would argue the DHCP loops are the result of the unstable 2.4Ghz band, not the cause.

@Mark112887 @Jon555 @joegreat @zoggy
There have been reports that the most recent release for AC3200 (382.50470) have solved the 2.4Ghz issue. A bunch of other devices also got 384.20942 updates too, so It is possible Asus solved the issue with new wireless drivers. However GPL archives are usually a few days behind, so we'll have to wait until those drop before Merlin can look at pulling the changes in.

 

[kfp]

Have it on stock asuswrt, on merlin with 68U and 86U even with new 68U beta from today

You're seeing it on 68U too? That's interesting, I would have thought this is a problem of running 32bit binaries on a 64bit processor which the 86U has.