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

[JSinFCVA]

Not sure if this would be of any help, but it might work. Just force the JFFS partition to be rebuilt at next boot.

 

[aex.perez]

View attachment 29484
here was my 7 day speed test W/ trendmicro on.(clients get around 900Mbps down and 40Mbps up). Model RT-AX88U
it starts with beta 3

then flawed beta 4

then beta 4b

Would it be safe to say your on cable? When you say trendmicro on, what parts of it? Also what scripts are running? I've had issues with Trendmicro features on, all off but Privacy accepted and with Skynet that impacted my results.

 

[Deleted member 22229]

Not sure if this would be of any help, but it might work. Just force the JFFS partition to be rebuilt at next boot.

He has already done that many times.

 

[SomeWhereOverTheRainBow]

Would it be safe to say your on cable? When you say trendmicro on, what parts of it? Also what scripts are running? I've had issues with Trendmicro features on, all off but Privacy accepted and with Skynet that impacted my results.

On home service I have cable with 1Gbps down and 35Mbps up plan . Scripts currently running are skynet, dnscrypt-proxy, diversion, connmon, spdmerlin,scmerlin,yazfi, scribe w/uiscribe, uidivstats, 2 active vpn client connections, and 2 openvpn site servers active as well routing traffic through vpn clients with x3mrouting. All of trend micro is active except QoS.

 

[Smokey613]

I had the unable to login via web also after going to beta4b. Everything else seemed to be working. I just let it run and went to bed. I got up this morning and logged right in, no issues.

I went back to 384.19 as the AiMesh 2.0 just does not work as well for me as the one in 384.19, I pretty much think this will be my place holder firmware for a long time. It just works without issues for me.

 

[aex.perez]

On home service I have cable with 1Gbps down and 35Mbps up plan . Scripts currently running are skynet, dnscrypt-proxy, diversion, connmon, spdmerlin,scmerlin,yazfi, scribe w/uiscribe, uidivstats, 2 active vpn client connections, and 2 openvpn site servers active as well routing traffic through vpn clients with x3mrouting. All of trend micro is active except QoS.

So I'm on 1Gb Fiber, had Skynet monitoring inbound and outbound. Then disabled it and enabled Trend AIProtect Two-Way IPS and Infected Device, thats the dip. Turned them both off, but not withdrawing from privacy, turned on Skynet but just inbound monitoring. So far my inbound and outbound are consistently in the ~920Mbps, off from the ~930Mbps when its disabled which matches what the PC gets. You're running a lot more than I am with VPN's and the scripts you have running.

 

[SomeWhereOverTheRainBow]

So I'm on 1Gb Fiber, had Skynet monitoring inbound and outbound. Then disabled it and enabled Trend AIProtect Two-Way IPS and Infected Device, thats the dip. Turned them both off, but not withdrawing from privacy, turned on Skynet but just inbound monitoring. So far my inbound and outbound are consistently in the ~920Mbps, off from the ~930Mbps when its disabled which matches what the PC gets. You're running a lot more than I am with VPN's and the scripts you have running.

I realize the router's self internet connection runs slower on speed test when trendmicro is on. Seems the other stuff such as scripts has no real effect. My wired clients get full speed despite trendmicro being active. The wireless internet throughput suffers a little. The main point that impacted connection was the faulty beta 4. The original beta 4 broke all internet throughput. Everything else seems to be working as expected. You are brave to shut off skynet... What would the world be like with out it.

 

[JSinFCVA]

He has already done that many times.

Oops. Sorry I misread the thread. Thought it was just multiple resets. I hope the hardware is not kaput. That would be unfortunate.

 

[kernol]

No, that was a different issue. The GPL code failed to properly compile the power management kernel module for the RT-AX86U, I had to fix the build recipe for it. Stock firmware wasn't affected because it built it from sources rather than prebuilt objects.

Many thanks for that ... and your continued efforts in a mighty complex multi-model project. {Thumbs-Up}

Sadly ... for now ... I must return to Asus Stock Version 3.0.0.4.386.41535 released for the RT-AX86U which, as I have said in other posts is the most stable stock release I have ever experienced for any model I have owned over the years.

With Merlinware 386.1_b4 I have the following insurmountable problems : -

• No Adaptive QoS without massive dcd taint errors caused by the version of Trend Micro included with Merlinware;
• Same applies to Guest Network 1 - get all the buggy errors on Merlin - but not with stock;
• Without Flex QoS the base Merlin Adaptive QoS gives very poor DSLReport scores [massive bufferbloat];
• Can't use CakeQoS on the RT-AX86U ... that addon needs a rewrite for this model [but frankly stock is good enough so CakeQoS not really necessary];

Asus stock Adaptive QoS works flawlessly without Flex QoS in the RT-AX86U - first to provide triple A+ scores on DSLReport.
With my limited fiber bandwidth of 100d/50u I need Adaptive QoS to work properly - and extending Guest 1 across the Aimesh Nodes is a very useful feature which I am exploiting [without buggy errors]

I really hope the quest to "normalise" your code across all supported models to ease the complex multi-model task will not serve to reduce all models to the lowest common denominator ... with high end models being compromised. I KNOW that is not your intention ... just an observation arising out of the apparent differences between stock released for the RT-AX86U and your beta releases.

 

[Kees17760]

I went back to 384.19 as the AiMesh 2.0 just does not work as well for me as the one in 384.19, I pretty much think this will be my place holder firmware for a long time. It just works without issues for me.

On my AC2900 this new build only generates more CPU heat and thus wear and tear, while i see no obvious advantage over 384.19.
384.19 is already very good at its job and it's hard to beat a winner ;-)

 

[SubElement]

Router has been up for 1day and 18 hours now now. Whatever was causing my router to crash and reboot is probably to do with either running scripts through amtm or possibly the SSD I had plugged in as storage for those scripts.

The investigation continues!

 

[beerglass007]

Managed to install OEM firmware and get TELNET [ Yuk ] enabled. I can now cli to the device and see the JFFS isn't mounted or something else

The host keys appear to be in /etc/dropbear and linked to /jffs/.ssh/ which it can't access

So now stuck as what is the problem

Any idea's please ?

Followed the guide this morning and still no SSH and the JFFS is unmounted state causing the problem

Brand new router dead that's it folks be careful no idea what caused that

1. Power off the router by pulling the power plug from the router and the AC wall plug too.
2. Power off the modem/ONT.
3. Remove all USB devices attached to the router. Do not re-insert them for this test.
4. Plug in the router's AC power plug to the wall outlet.
5. Make sure the routers power switch is in the off position.
6. Plug in the power plug to the router.
7. Hold down the WPS button and keep holding it until the router reboots/turns off while you turn the router on using the power button.
8. When the router boots up, it should be in the Setup Wizard. If it isn't, repeat the steps above until it is.
9. Use the following URL to bypass the setup wizard and go straight to the factory reset part of the GUI.
10. http://router.asus.com/Advanced_SettingBackup_Content.asp
11. Check the 'Initialize all the settings...' box.
12. Click on the 'Restore' button beside it.
13. After the router reboots, use the same URL above to flash the firmware you have downloaded and verified at the beginning of this post.
14. After the router reboots, use the same URL to repeat steps 11, 12 and 13 above.
15. Plug in the modem/ONT and connect the Ethernet cable to the router's WAN port.
16. After the router reboots use the Setup Wizard to connect to your ISP and secure your router.
    1. Use a new (never before used) SSID that is 8 alphanumeric characters with no spaces, punctuation, smiley faces or other special characters.
    2. Use all-new (never before used) passwords that are 16 alphanumeric characters with no spaces, punctuation, smiley faces or other special characters.
    3. Use a new (never before used) username that is 8 alphanumeric characters with no spaces, punctuation, smiley faces or other special characters.
    4. Do not change any other settings or enable any other features.
17. Do not insert any USB devices.
18. After the router has been running for 15 minutes or more, reboot it once more.

 

[Unisoft]

Many thanks for that ... and your continued efforts in a mighty complex multi-model project. {Thumbs-Up}

Sadly ... for now ... I must return to Asus Stock Version 3.0.0.4.386.41535 released for the RT-AX86U which, as I have said in other posts is the most stable stock release I have ever experienced for any model I have owned over the years.

With Merlinware 386.1_b4 I have the following insurmountable problems : -

• No Adaptive QoS without massive dcd taint errors caused by the version of Trend Micro included with Merlinware;
• Same applies to Guest Network 1 - get all the buggy errors on Merlin - but not with stock;
• Without Flex QoS the base Merlin Adaptive QoS gives very poor DSLReport scores [massive bufferbloat];
• Can't use CakeQoS on the RT-AX86U ... that addon needs a rewrite for this model [but frankly stock is good enough so CakeQoS not really necessary];

Asus stock Adaptive QoS works flawlessly without Flex QoS in the RT-AX86U - first to provide triple A+ scores on DSLReport.
With my limited fiber bandwidth of 100d/50u I need Adaptive QoS to work properly - and extending Guest 1 across the Aimesh Nodes is a very useful feature which I am exploiting [without buggy errors]

I really hope the quest to "normalise" your code across all supported models to ease the complex multi-model task will not serve to reduce all models to the lowest common denominator ... with high end models being compromised. I KNOW that is not your intention ... just an observation arising out of the apparent differences between stock released for the RT-AX86U and your beta releases.

For comparison, despite initial problems with ALPHAs and early beta, Im pretty much OK except for a device icon issue in the web pages where they randomly dont show at their correct size and the layout screws up the gui tables.

MY QOS works fine with Beta4B. I use DLSReports and get A+ on everything. I put Gaming as low because I don't need it.

I have in this order of priority:

• Work from Home
• Web Surfing
• Audio and Video Streaming
• File Transferring
• Others
• Gaming
• Learn from Home

with manual upload and download values set and GEFROCE NOW off.

I used to use FRESH QOS, but don't now.

I have TWO AX86U routers - one at each address. One is 383mbps down and 37 upstream (average it always hits) - and the other is 219mbps and 21.
Both are cable broadband into a Virgin Media router set as Modem Mode only.

Both are fine with QOS for me with bufferbloat?

 

[Andy1932]

Many thanks for that ... and your continued efforts in a mighty complex multi-model project. {Thumbs-Up}

Sadly ... for now ... I must return to Asus Stock Version 3.0.0.4.386.41535 released for the RT-AX86U which, as I have said in other posts is the most stable stock release I have ever experienced for any model I have owned over the years.

With Merlinware 386.1_b4 I have the following insurmountable problems : -

• No Adaptive QoS without massive dcd taint errors caused by the version of Trend Micro included with Merlinware;
• Same applies to Guest Network 1 - get all the buggy errors on Merlin - but not with stock;
• Without Flex QoS the base Merlin Adaptive QoS gives very poor DSLReport scores [massive bufferbloat];
• Can't use CakeQoS on the RT-AX86U ... that addon needs a rewrite for this model [but frankly stock is good enough so CakeQoS not really necessary];

Asus stock Adaptive QoS works flawlessly without Flex QoS in the RT-AX86U - first to provide triple A+ scores on DSLReport.
With my limited fiber bandwidth of 100d/50u I need Adaptive QoS to work properly - and extending Guest 1 across the Aimesh Nodes is a very useful feature which I am exploiting [without buggy errors]

I really hope the quest to "normalise" your code across all supported models to ease the complex multi-model task will not serve to reduce all models to the lowest common denominator ... with high end models being compromised. I KNOW that is not your intention ... just an observation arising out of the apparent differences between stock released for the RT-AX86U and your beta releases.

I can't speak for its effectiveness, but the AX86u is now supported by CakeQOS.

 

[hancox]

I can't speak for its effectiveness, but the AX86u is now supported by CakeQOS.

Yeah - i was scratching my head here, too - i have CakeQOS running without issue on an AX-86U with b4.

 

[brianward317]

AX-88U has been up for 4 days now without any issues since the latest firmware was published beta4b. Running most of the scripts/nextdns CLI/trendmicro/QOS Nice work @RMerlin

Other two aimesh nodes are running smooth as well.

 

[ajp2k14]

@RMerlin I'm seeing half of the normal OpenVPN client speeds with my AC86U on 386.1 B4b. Any idea what could be causing this and what I could do to troubleshoot?

EDIT: Also I think the cpu usage is higher on 386.1B when using openvpn...

 

[aex.perez]

I realize the router's self internet connection runs slower on speed test when trendmicro is on. Seems the other stuff such as scripts has no real effect. My wired clients get full speed despite trendmicro being active. The wireless internet throughput suffers a little. The main point that impacted connection was the faulty beta 4. The original beta 4 broke all internet throughput. Everything else seems to be working as expected. You are brave to shut off skynet... What would the world be like with out it.

Brave for about the time it took to run a few bandwidth tests, ended up with just inbound monitoring went from ~920/~600 to ~920/~920, Trendmicro hit for me cuts down and up speeds by 1/2. Going to set Skynet to monitor all, with just the Trendmicro privacy on to see if I can trigger the drop that happened the other day where I actually had to withdraw, even with all the features off, to get my performance back.

 

[Psu2ski32]

Hi all- Flashed to 386 beta4 on AC5300 and AC3100 last night. Pretty standard set up with no scripts, etc. Overall working pretty well. Did a bit of playing and was able to bind clients to a couple of my AiMesh Nodes. That said, had a quick question. I am getting faster speeds on my laptop when running speed test compared to speed test on main AC5300 router. I do have 1Gig Speed coming into the Cable Modem. Any ideas why this would reflect this way? I played a bit with AiProtection and withdrawing privacy notice but that didnt seem to change anything.

Router Speeds:

Laptop (Apple Device) Speeds:

 

[Makaveli]

Hi all- Flashed to 386 beta4 on AC5300 and AC3100 last night. Pretty standard set up with no scripts, etc. Overall working pretty well. Did a bit of playing and was able to bind clients to a couple of my AiMesh Nodes. That said, had a quick question. I am getting faster speeds on my laptop when running speed test compared to speed test on main AC5300 router. I do have 1Gig Speed coming into the Cable Modem. Any ideas why this would reflect this way? I played a bit with AiProtection and withdrawing privacy notice but that didnt seem to change anything.

Router Speeds:

View attachment 29516

Laptop (Apple Device) Speeds:

View attachment 29517

The speed test run directly on the router will be handicapped by the router cpu.If you are getting full speed using the speedtest app then you are good.

I see alot of people complaining about this.

Speedtest app > internal router test.