Beta Asuswrt-Merlin 386.1 Beta (stage 2) is now available

[dev_null]

I have an interesting issue. My AX86U running 386.1b5 is erasing cron entries without any other symptoms. I only noticed this because I didn't get an expected overnight report. This happened once on b4 but the device was brand new to me and I assumed I misconfigured something.

These cron jobs are added through services-start. The device is an AX86U running 386.1b5 with about 35 devices connected. Temps are in the 50-60C range. The device uptime does not change - so it's not the device rebooting. There is nothing I can see in the logs. This has not happened on my other devices even on the 386.1 beta software (AC66U_B1/AC68U).

Anyone else notice disappearing cron jobs? Any potential solutions?

EDIT: narrowed the time down based on when NTPmerlin stopped recording. The logs around that time are below.

Spoiler

Jan 27 21:15:01 RT-AX86U-C328 (ChkWAN.sh): 10915 v1.11 Monitoring WAN connection using 2 target PING hosts (72.251.239.206 162.249.171.198) (Tries=3)
Jan 27 21:15:01 RT-AX86U-C328 (ChkWAN.sh): 10915 Monitoring pass 1 out of 3
Jan 27 21:15:01 RT-AX86U-C328 (ChkWAN.sh): 10915 Monitoring WAN connection OK.....(Successful ping to '72.251.239.206'); Terminating due to ACTIVE cron schedule
Jan 27 21:21:28 RT-AX86U-C328 kernel: Init chrdev /dev/idp with major 190
Jan 27 21:21:28 RT-AX86U-C328 kernel: tdts: tcp_conn_max = 8000
Jan 27 21:21:28 RT-AX86U-C328 kernel: tdts: tcp_conn_timeout = 300 sec
Jan 27 21:21:30 RT-AX86U-C328 kernel: SHN Release Version: 2.0.1 4b635f32
Jan 27 21:21:30 RT-AX86U-C328 kernel: UDB Core Version: 0.2.18
Jan 27 21:21:30 RT-AX86U-C328 kernel: Init chrdev /dev/idpfw with major 191
Jan 27 21:21:30 RT-AX86U-C328 kernel: IDPfw: flush fc
Jan 27 21:21:30 RT-AX86U-C328 kernel: IDPfw: IDPfw is ready
Jan 27 21:21:30 RT-AX86U-C328 kernel: sizeof forward pkt param = 280
Jan 27 21:21:30 RT-AX86U-C328 BWDPI: fun bitmap = 3
Jan 27 21:21:30 RT-AX86U-C328 kernel: CONSOLE: 034460.978 wl1.0: wlc_send_bar: for b0:ca:68:15:b2:5a seq 0x12 tid 1
Jan 27 21:21:41 RT-AX86U-C328 BWDPI: force to flush flowcache entries
Jan 27 21:21:41 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:21:41 RT-AX86U-C328 kernel: mod epilog takes 0 jiffies
Jan 27 21:21:41 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:21:41 RT-AX86U-C328 kernel: Exit chrdev /dev/idpfw with major 191
Jan 27 21:21:41 RT-AX86U-C328 kernel: Exit chrdev /dev/idp with major 190
Jan 27 21:21:41 RT-AX86U-C328 custom_script: Running /jffs/scripts/nat-start
Jan 27 21:21:41 RT-AX86U-C328 custom_script: Running /jffs/scripts/firewall-start (args: eth0)
Jan 27 21:21:41 RT-AX86U-C328 CakeQOS-Merlin: Starting - ( 105Mbit | 5Mbit | diffserv4 | docsis wash | docsis ack-filter )
Jan 27 21:21:41 RT-AX86U-C328 kernel: Cpuidle Host Clock divider is disabled
Jan 27 21:21:42 RT-AX86U-C328 kernel: CONSOLE: 034472.863 wl1.0: wlc_send_bar: for 7c:a1:ae:c4:42:b5 seq 0x17 tid 1
Jan 27 21:22:33 RT-AX86U-C328 kernel: Init chrdev /dev/idp with major 190
Jan 27 21:22:33 RT-AX86U-C328 kernel: tdts: tcp_conn_max = 8000
Jan 27 21:22:33 RT-AX86U-C328 kernel: tdts: tcp_conn_timeout = 300 sec
Jan 27 21:22:35 RT-AX86U-C328 kernel: SHN Release Version: 2.0.1 4b635f32
Jan 27 21:22:35 RT-AX86U-C328 kernel: UDB Core Version: 0.2.18
Jan 27 21:22:35 RT-AX86U-C328 kernel: Init chrdev /dev/idpfw with major 191
Jan 27 21:22:35 RT-AX86U-C328 kernel: IDPfw: flush fc
Jan 27 21:22:35 RT-AX86U-C328 kernel: IDPfw: IDPfw is ready
Jan 27 21:22:35 RT-AX86U-C328 kernel: sizeof forward pkt param = 280
Jan 27 21:22:35 RT-AX86U-C328 BWDPI: fun bitmap = 3
Jan 27 21:22:49 RT-AX86U-C328 BWDPI: force to flush flowcache entries
Jan 27 21:22:49 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:22:49 RT-AX86U-C328 kernel: mod epilog takes 0 jiffies
Jan 27 21:22:49 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:22:50 RT-AX86U-C328 kernel: Exit chrdev /dev/idpfw with major 191
Jan 27 21:22:50 RT-AX86U-C328 kernel: Exit chrdev /dev/idp with major 190
Jan 27 21:22:50 RT-AX86U-C328 custom_script: Running /jffs/scripts/nat-start
Jan 27 21:22:50 RT-AX86U-C328 custom_script: Running /jffs/scripts/firewall-start (args: eth0)
Jan 27 21:22:50 RT-AX86U-C328 CakeQOS-Merlin: Starting - ( 105Mbit | 5Mbit | diffserv4 | docsis wash | docsis ack-filter )
Jan 27 21:22:50 RT-AX86U-C328 kernel: Cpuidle Host Clock divider is disabled
Jan 27 21:23:15 RT-AX86U-C328 kernel: Init chrdev /dev/idp with major 190
Jan 27 21:23:15 RT-AX86U-C328 kernel: tdts: tcp_conn_max = 8000
Jan 27 21:23:15 RT-AX86U-C328 kernel: tdts: tcp_conn_timeout = 300 sec
Jan 27 21:23:17 RT-AX86U-C328 kernel: SHN Release Version: 2.0.1 4b635f32
Jan 27 21:23:17 RT-AX86U-C328 kernel: UDB Core Version: 0.2.18
Jan 27 21:23:17 RT-AX86U-C328 kernel: Init chrdev /dev/idpfw with major 191
Jan 27 21:23:17 RT-AX86U-C328 kernel: IDPfw: flush fc
Jan 27 21:23:17 RT-AX86U-C328 kernel: IDPfw: IDPfw is ready
Jan 27 21:23:17 RT-AX86U-C328 kernel: sizeof forward pkt param = 280
Jan 27 21:23:17 RT-AX86U-C328 BWDPI: fun bitmap = 3
Jan 27 21:25:38 RT-AX86U-C328 BWDPI: force to flush flowcache entries
Jan 27 21:25:38 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:25:38 RT-AX86U-C328 kernel: mod epilog takes 0 jiffies
Jan 27 21:25:38 RT-AX86U-C328 kernel: IDPfw: Exit IDPfw
Jan 27 21:25:38 RT-AX86U-C328 kernel: Exit chrdev /dev/idpfw with major 191
Jan 27 21:25:38 RT-AX86U-C328 kernel: Exit chrdev /dev/idp with major 190
Jan 27 21:25:38 RT-AX86U-C328 custom_script: Running /jffs/scripts/nat-start
Jan 27 21:25:38 RT-AX86U-C328 custom_script: Running /jffs/scripts/firewall-start (args: eth0)
Jan 27 21:25:38 RT-AX86U-C328 CakeQOS-Merlin: Starting - ( 105Mbit | 5Mbit | diffserv4 | docsis wash | docsis ack-filter )
Jan 27 21:25:38 RT-AX86U-C328 kernel: Cpuidle Host Clock divider is disabled
Jan 27 21:27:19 RT-AX86U-C328 hostapd: eth6: STA fc:f8:ae:be:7a:c1 IEEE 802.11: associated
Jan 27 21:27:19 RT-AX86U-C328 kernel: br0: received packet on eth6.0 with own address as source address
Jan 27 21:27:19 RT-AX86U-C328 kernel: CFG80211-ERROR) wl_cfg80211_change_station : WLC_SCB_AUTHORIZE sta_flags_mask not set
Jan 27 21:27:19 RT-AX86U-C328 hostapd: eth6: STA fc:f8:ae:be:7a:c1 RADIUS: starting accounting session BCFC4DA2ED18043A
Jan 27 21:27:19 RT-AX86U-C328 hostapd: eth6: STA fc:f8:ae:be:7a:c1 WPA: pairwise key handshake completed (RSN)
Jan 27 21:27:52 RT-AX86U-C328 kernel: Init chrdev /dev/idp with major 190
Jan 27 21:27:52 RT-AX86U-C328 kernel: tdts: tcp_conn_max = 8000
Jan 27 21:27:52 RT-AX86U-C328 kernel: tdts: tcp_conn_timeout = 300 sec
Jan 27 21:27:53 RT-AX86U-C328 kernel: SHN Release Version: 2.0.1 4b635f32
Jan 27 21:27:53 RT-AX86U-C328 kernel: UDB Core Version: 0.2.18
Jan 27 21:27:53 RT-AX86U-C328 kernel: Init chrdev /dev/idpfw with major 191
Jan 27 21:27:53 RT-AX86U-C328 kernel: IDPfw: flush fc
Jan 27 21:27:53 RT-AX86U-C328 kernel: IDPfw: IDPfw is ready
Jan 27 21:27:53 RT-AX86U-C328 kernel: sizeof forward pkt param = 280

 

[mromero]

with all of these "my router is running hotter" posts, I wonder how many people have had a look at the venting on the backs of their routers?
I'm betting that some people may benefit from a small amount of dust removal maintenance from the vents with a vacuum and possibly some compressed air.
There are options to add fans for active airflow to compliment the passive cooling if people deem it necessary...but even 80-85 is well within design spec for the machine to work correctly...and there are places in the world where a 30C room is uncomfortably warm, so even with unblocked vents and added airflow, it still might not cool off much
<shrug> take a breath and check the situation beyond what the router is reporting, people @RMerlin is trying to get 386.1 to release - chasing your thermometers is a bit counterproductive this close to the finish line, wouldn't you agree?

Router is six moths old and clean as a whistle. Purchasing and attaching fans is NOT why I purchased this router.

Past tear downs by Mr. Higgins have shown where certain Asus routers were designed with internal fans and this feature never implemented. Our Cisco routers are built like tanks and have internal fans that are cleaned out once a year.

FWIW 30C is NOT uncomfortable with the breeze from the Caribbean Sea blowing through.

Today's temperature report. Room temp 22.222°C

Legend: 2.4 GHz - 5 GHz - CPU

Current Temperatures: 47 °C - 53 °C - 82 °C

Just went back to 384.19, and here is your temperature report; reduced cooking the CPU by 8 °C

Legend: 2.4 GHz - 5 GHz - CPU
Current Temperatures: 47 °C - 53 °C - 74 °C

 

[Smokey613]

I was wondering, are there any advantages upgrading to this latest Merlin firmware since I am running in AP mode? Are there significant improvements on the wifi that I would benefit from?

 

[Slawek P]

- I have no longer seen phantom (disconnected) devices in network map, so in this respect this beta seems like having improved since bet4b.

Interesting - I still see phantom devices on mine. If something was conntected via WiFi to AiMesh node then I often see it ghosting as an ethernet device when it is disabled.

 

[clvk07]

Check your system log, it will tell you if/why dropbear isn't running.

After upgrade I Factory reseat the router and rebooted via console. Once I rebooted the router with the power switch it worked. All good now.

 

[maxbraketorque]

Router is six moths old and clean as a whistle. Purchasing and attaching fans is NOT why I purchased this router.

Past tear downs by Mr. Higgins have shown where certain Asus routers were designed with internal fans and this feature never implemented. Our Cisco routers are built like tanks and have internal fans that are cleaned out once a year.

FWIW 30C is NOT uncomfortable with the breeze from the Caribbean Sea blowing through.

Today's temperature report. Room temp 22.222°C

Legend: 2.4 GHz - 5 GHz - CPU

Current Temperatures: 47 °C - 53 °C - 82 °C

Just went back to 384.19, and here is your temperature report; reduced cooking the CPU by 8 °C

Legend: 2.4 GHz - 5 GHz - CPU
Current Temperatures: 47 °C - 53 °C - 74 °C

My AC86Us are sitting around 80-81C with 386.1b5, which is not bad, but I decided to order a fan. Applying fresh thermal compound between the CPU and the heat sink will also significantly reduce temperature.

 

[RMerlin]

So I have to reduce the DHCP table to max 64 entries. Any change that the final release will allow 128 entries again? Maybe a stupid question, but I didn't notice this decrease in nr of entries in the change log (or maybe I missed it).

No. 128 was never attainable on the majority of routers because there isn`t enough nvram space to store 128 of them.

 

[GHog]

Just an observation.
I have not upgraded to B5 yet, but I notice that there are lots of strange issues that might be contributed to users that use VOIP boxes. For example, OOMA wants you to connect in series with router ,IF, you cannot connect to separate lan port on your providers box. So here is what I have found after a few years of using devices connected in series with ASUS wireless routers. DO NOT DO THAT. What I have done for myself and many others is connect the VOIP box to a lan port coming out of your service providers interface. If your device has only one lan port, go buy the best 4 port switch you can and connect it to single output port of your providers box and then your ASUS device to lan port 1 and your VOIP to lan port 2. You will not be able to see directly into your VOIP device through your router now. No big deal. Unplug what ever you have on router lan port and connect it to lan port on VOIP to configure it if needed. You can still go to your VOIP web page to do what ever you like when you are connected through ASUS router. Once I started doing this, my router stopped having strange issues like I see posted. ASUS and other wireless routers really like a clean dedicated connection to your providers device. If your providers box does have multiple lan ports, just do what we all do and turn wifi radio off in the providers box and connect your router.
I know for most this is obvious, but not for many out there suffering with issues.
I have been using Draytek Vigor 2762, 2862B, line connections devices and they are stable and work better than anything I have tried for DSL service. For single port fiber line connection boxes, I just get a high speed 4 port switch.
Hope this helps

 

[sentinelvdx]

Beta 5 is causing CPU spikes or stays high on my AC88U, but not seeing anything on syslog.
CPU Temp 87c (22c ambient)

 

[FTC]

Interesting - I still see phantom devices on mine. If something was conntected via WiFi to AiMesh node then I often see it ghosting as an ethernet device when it is disabled.

Could be very much that those have just *not yet* appeared in my installation (only 40 hours running now) or that you are not waiting enough ? I menant that whenever I see a device that is not connected it usually disappears in less than 1 minute. In previous betas I used to see consistently some devices that had been disconnected for hours.

By the way, is your aimesh node also in beta5 ?

EDIT : After almost 60 hours working, I have again some ghost devices, so this one has not been fixed.

 

[John Fitzgerald]

Beta 5 is causing CPU spikes or stays high on my AC88U, but not seeing anything on syslog.
CPU Temp 87c (22c ambient)

2 things to try:
Dirty power off, pull the barrel plug wait 1min, then plug back in.
Force reinstall all scripts, AMTM...uu, Skynet option 10, sub option 3, Diversion...uu, FlexQOS you will have look up.
Unbound...option 1.

 

[spedwards]

I was wondering, are there any advantages upgrading to this latest Merlin firmware since I am running in AP mode? Are there significant improvements on the wifi that I would benefit from?

I think it depends upon your Asus model and which FW you are migrating from. In short YMMV.

With my standalone AX88U the 2.4Ghz WiFi is immensly improved both on stability and range after coming from 384.19 to 386.1_beta5.

For example, my two furthest Ring cameras both now report a strong WiFi signal and are completely stable. These two were intermittently dropping off WiFi several times each day due to range issues.
The 2.4Ghz WiFi issues with 384.19 have been discussed elsewhere in the forum. It wasn't present in some earlier FW, so you may not be experiencing it yourself.

The 5Ghz WiFi performance doesn't seem to have changed at all for my setup. It was working very well and it has stayed working very well.

The GUI appears snappier and there are new additions. But I don't know what would interest you as an AP except the WiFi update.

 

[spedwards]

No. 128 was never attainable on the majority of routers because there isn`t enough nvram space to store 128 of them.

Our home IOT collection has pushed our current content to 48 DHCP clients. Looks Like it'll hit 64 before too long at the rate we are adding lightbulbs, mains switches and other gadgets. Can you run a second DHCP server on AIMESH nodes to increase the overall capacity?

How large is the table in Asus FW for the AX88U?

 

[sentinelvdx]

2 things to try:
Dirty power off, pull the barrel plug wait 1min, then plug back in.
Force reinstall all scripts, AMTM...uu, Skynet option 10, sub option 3, Diversion...uu, FlexQOS you will have look up.
Unbound...option 1.

I'll give a try later. Right now I've just did a dirty power off.
Anyway it's a beta, things like this are expected. This weekend I'll probably do a hard reset and reconfigure/reinstall everything back

Thanks

 

[heysoundude]

Router is six moths old and clean as a whistle. Purchasing and attaching fans is NOT why I purchased this router.

Past tear downs by Mr. Higgins have shown where certain Asus routers were designed with internal fans and this feature never implemented. Our Cisco routers are built like tanks and have internal fans that are cleaned out once a year.

FWIW 30C is NOT uncomfortable with the breeze from the Caribbean Sea blowing through.

Today's temperature report. Room temp 22.222°C

Legend: 2.4 GHz - 5 GHz - CPU

Current Temperatures: 47 °C - 53 °C - 82 °C

Just went back to 384.19, and here is your temperature report; reduced cooking the CPU by 8 °C

Legend: 2.4 GHz - 5 GHz - CPU
Current Temperatures: 47 °C - 53 °C - 74 °C

turn it away from the beach view? I run a bit hot with bikinis around too ;-p

 

[CaptnDanLKW]

Running 2xAC86Us in AI Mesh. I first made the jump from 384_19 to 386.1_b4b. a few weeks ago. Upgraded to beta 5 the day it came out.
I don't use VPN, nor Skynet or any QOS.
I do have the scribe, uiScribe, scmerlin and conmon addons.
I did NOT reset my configuration

I don't have a strong opinion either way about the CPU temperature increase that has been debated but my CPU Temp range either idles around 84C or around 90C. Checking htop and other metrics, I don't see why its so high with light load, nor can I tell why it picks these 2 temps. It still runs hotter than 384_19, but no reason for me to go back. FWIW, I checked the pwr command and the cpu wait is enabled.

AI Mesh with wired backhaul and SmartConnect is working phenomenal.

The built-in speed test is reliable and on par with an official app on my desktop. (600M Down/10M Up)

Also noticed, my syslog has never been QUIETER. (albeit, I enabled kernel, and ioctl syslog filters).

While its all running perfect, I'll probably nuke and re-do the config at some point in the future, however I'm very happy I made the jump.

 

[John Fitzgerald]

I'll give a try later. Right now I've just did a dirty power off.
Anyway it's a beta, things like this are expected. This weekend I'll probably do a hard reset and reconfigure/reinstall everything back

Thanks

When you go the long road version, be sure to wipe the JFFS partition and reformat the USB drive in a PC to NTFS or something else (just don't use an EXT derivation), so when you port it back and let AMTM format it to EXT4 with Journaling 2GB Swap file, it's as clean as it can be.

Making the point for anyone else that needs the info.

 

[fg64]

Dirty upgrade from b3 to b5. All good so far. No issues detected. Thanks again for all the hard work.

 

[tdhite]

Our home IOT collection has pushed our current content to 48 DHCP clients. Looks Like it'll hit 64 before too long at the rate we are adding lightbulbs, mains switches and other gadgets. Can you run a second DHCP server on AIMESH nodes to increase the overall capacity?

How large is the table in Asus FW for the AX88U?

Frankly it's as large as JFFS or your USB drive if you point at that for the address management. For example, you can manage the addresses in /jffs/configs/dnsmasq.conf.add directly or elsewhere. For example, in /jffs/configs/dnsmasq.conf.add, add something like:

dhcp-hostsfile=/jffs/configs/dnsmasq.hosts.iot

in /jffs/configs/dnsmasq.hosts.iot put all your entries in the form, one per line:

MACADDR,IPADDR,DNSNAME

Example line:

12:23:34:45:56:67,172.16.10.233,my-iot-device1.mydomain.com

or the like. No need for nvram at all when done that way.

 

[AntonK]

Beta 5 is causing CPU spikes or stays high on my AC88U, but not seeing anything on syslog.
CPU Temp 87c (22c ambient)

My 88U is not showing the behaviors yours is.