[Release] Asuswrt-Merlin 384.12 is now available

[Asad Ali]

After playing and playing: I have found that on my AX88U, if I enable the ntpd-server and set it to intercept ntp traffic, and then reboot, the boot process is stalled, by the ntp update not happening yet. The system gets an IP and all that but has no time set at this point. If I wait it will set the time, eventually, and any time critical features seem to restart ok. Something is broken with the new ntpd-server feature on this model router.

Using any scripts which can mess up with ntp ready nvram variable? Try again after disabling scripts execution on the router.

 

[Swistheater]

After playing and playing: I have found that on my AX88U, if I enable the ntpd-server and set it to intercept ntp traffic, and then reboot, the boot process is stalled, by the ntp update not happening yet. The system gets an IP and all that but has no time set at this point. If I wait it will set the time, eventually, and any time critical features seem to restart ok. Something is broken with the new ntpd-server feature on this model router.

Using any scripts which can mess up with ntp ready nvram variable? Try again after disabling scripts execution on the router.

i concur with asad ali about messing with the ntp_ready variable , it will force the ntpd to not report accurately.
like with this https://gist.github.com/cynicastic/ced78fac27de4394b67977802d76c0d9 well thought out script, it will allow for a access point or a router that has another local time source to be set correctly for that instance, but this is not effective way to work if you are using DoT on that router that the script is ran on. there are also other numerous variables that rely on ntp_ready function as well.

alternatively, I propose finding out what custom scripts run around the time your router says time has been synced and add a script to run with it

logger -t "$(basename $0)" "waiting for Network..."
until ping -nq -c3 1.1.1.1; do
   sleep 1
done
service restart_ntpd

this will allow it to restart ntpd once internet connection has been established, obviously your router is doing its time sync well before you have fully established a wan connection. or either you may need to look for port configurations if you are having a double nat issue.

One option is if you run disk check with amtm, you can always place this to occur in the pre-mount right after diskcheck has occurred.

 

[Swistheater]

After playing and playing: I have found that on my AX88U, if I enable the ntpd-server and set it to intercept ntp traffic, and then reboot, the boot process is stalled, by the ntp update not happening yet. The system gets an IP and all that but has no time set at this point. If I wait it will set the time, eventually, and any time critical features seem to restart ok. Something is broken with the new ntpd-server feature on this model router.

are you back to using a static IP? I noticed when i switch from static to dynamic, I had to do a factory reset because this issue started occurring to me as well.

try completely unplugging every thing. (restart your modem as well with no ethernet cable plugged into it). wait for modem to fully re-establish connection, then plug your router back in and reconnect cables and power the router back on. see if you are still having this behavior, if so you may need to factory reset.

 

[MarkyMarkMark]

First time running Asuswrt-Merlin but kind of familiar since I ran johnsfork on RT-N66U for a long time.
So I upgraded to a Arm based router, the RT-AC68U and this is my first go with rt-Merlin 384.12.

After getting everything setup, my first question is regarding the USB display on network map. Neither my usb sticks (1 ext4 , 1 ext2) say mounted? There is a blank box under the disk name.

I'm running Diversion off of the ext2 stick and when I SSH into the box they both come up /dev/sda and /dev/sdb, via "mount" command line. Just wondering if this is a graphical issue or if I somehow haven't formatted the USB sticks correctly?

The stick running Diversion was used in my previous RT-N66U and worked fine. Even installing Diversion I was asked if I wanted to restore setup that was already on the stick, which I declined for a fresh install.

 

[RMerlin]

Hello, after updating the RT-AC68U router from the firmware version 384.10 I have a problem with the VPN server. In the 384.9 firmware version, the VPN server operates correctly. Attached screenshot. What could be the reason ? Thank you for your help.

Check your System Log, it will tell you what is wrong.

 

[RMerlin]

Update over b2 worked without any bigger issues. The only minor problem I encountered, I lost all of my QoS setting. All settings where set back to their default values. This happened on both routers after updating.

Known issue that's been there for months, Asus only recently fixed it on their end, but haven't released any new firmware since fixing it.

After playing and playing: I have found that on my AX88U, if I enable the ntpd-server and set it to intercept ntp traffic, and then reboot, the boot process is stalled, by the ntp update not happening yet. The system gets an IP and all that but has no time set at this point. If I wait it will set the time, eventually, and any time critical features seem to restart ok. Something is broken with the new ntpd-server feature on this model router.

Works for me.

The main difference when enabling that option is a new firewall rule is created to redirect traffic. If you use any customized firewall rules, review your changes.

 

[L&LD]

In my opinion router reset is needed only in very few situations when something went completely wrong and the router is not communicating and is not accessible. I see a trend here on SNB to recommend a reset for almost everything without even trying to help the user by asking the right questions. Just reset and start over is a wrong recommendation without identifying the issue in the first place. It sounds to me like "buddy, we are pros here, don't bother us with your stuff, just start over". I see reset recommendations in every thread now, no matter what the issue is.

If that is what you are getting out of it, you are reading the posts wrong.

Nobody knows the state a users router is, nor what steps were taken to get there. The user themselves couldn't tell you with any certainty in most cases how the router got there except in broad terms. Even a brand new router is not as in a pristine state as most users believe they are. Today's returns, exchanges, and even manufacturers own testing can leave a router with inter-firmware glitches that given a certain configuration, will show up as insolvable bugs or even hint at hardware failure when they're not.

What does jump out at me is when all other reports of a similar configuration just work, yet a particular unit doesn't. That indicates that a router is in a very unstable state for any of the possible reasons above.

The fastest way to get out of that state is a full reset to factory defaults and progressively more 'nuclear' attacks on it, as necessary. The reality is that a basic reset to factory defaults is only a few minutes saved vs. a full Nuclear Reset of the router itself. Meaning, it is usually faster to perform the latter than to perform both (as is usually seen in users posts).

When I go to a customers site and they just want their network running as optimally as possible with the equipment they have, a Nuclear Reset is the most professional option I can offer them (but, I don't call it that, of course). This is not a 'don't bother me' type of response! My proof is that I very rarely get called back to those customers. And when I do, the fix is usually a 15-minute process, at most.

Of course, I spend 10 to 30 minutes or more (when customers try to explain things to me) to be sure I understand the issues I'm trying to fix before recommending a full reset, but usually, I find that is wasted time because almost any fixes I try at that point do not stick (at least not permanently) and rebooting the router isn't something the customer wants to do either, indefinitely.

Speaking specifically of Asus routers powered by RMerlin firmware or not, a reset is necessary when the router isn't responding as it should (not necessarily as the user fancies, but when it doesn't respond within it's designed specs and firmware capabilities) and a reasonable search for fixes comes up empty-handed (or, the fixes presented simply don't work on that router/network).

There are a few posts here on these forums where a full reset brought a router back to life when all other indications seemed otherwise.

https://github.com/RMerl/asuswrt-merlin/wiki/Installation

From the link above.

• If something looks weird, don't waste too much time: save your settings, reset to factory default, reconfigure the basics, and see if the issue is resolved. If not, you can always restore your saved settings and do some more advanced troubleshooting.
• It is not recommended to restore settings saved under a different firmware version. It might work, but there is no guarantee.

 

[cmkelley]

i concur with asad ali about messing with the ntp_ready variable , it will force the ntpd to not report accurately.
like with this https://gist.github.com/cynicastic/ced78fac27de4394b67977802d76c0d9 well thought out script, it will allow for a access point or a router that has another local time source to be set correctly for that instance, but this is not effective way to work if you are using DoT on that router that the script is ran on. there are also other numerous variables that rely on ntp_ready function as well.

I finally realized why others have problems with this and I don't. The local time source I point to is my primary time source (a gps-enabled Raspberry Pi), so in _my_ case, once it can reach my local time source, it's fine to set ntp_ready because ntp will be able to reach that time source.

I'll update the gist to discuss this.

 

[QuikSilver]

It sounds to me like "buddy, we are pros here, don't bother us with your stuff, just start over".

I can see how some get that impression seeing reset suggestions but I’ve been following along with the threads here long enough to see where it has helped a few people. Resets aren’t for everyone and I don’t think any of the “pros” here suggest that just to ignore you but I do think they want people to read and exhaust all options BEFORE stating that the firmware/script is to blame when others clearly don’t have that same issue. A lot of times you will see others have the same issue and may be able to word the question better than someone else AFTER someone previously stated to reset it.




Sent from my iPhone using Tapatalk

 

[Val D.]

...full Nuclear Reset...

This so called Nuclear Reset is the ultimate waste of time. I know what's inside the routers at component level and how it works as electronics. Half of the entire "procedure" is completely unnecessary and does the same thing initiated in a different way. Also recommendations to wait 5 min, 10 min, 1 hour and so on are waste of time steps. I can guarantee no electrons are flowing in any direction inside the router 60 seconds after power cord is unplugged. If a router doesn't perform normally after normal reset procedure, then this device has a hardware issue and chances to fix it after Nuclear Reset, Deep Freezing or Casting Spells are minimal. The device just has to be replaced. Routers fail for different reasons and have to be replaced like any other electronic device.

 

[L&LD]

This so called Nuclear Reset is the ultimate waste of time. I know what's inside the routers at component level and how it works as electronics. Half of the entire "procedure" is completely unnecessary and does the same thing initiated in a different way. Also recommendations to wait 5 min, 10 min, 1 hour and so on are waste of time steps. I can guarantee no electrons are flowing in any direction inside the router 60 seconds after power cord is unplugged. If a router doesn't perform normally after normal reset procedure, then this device has a hardware issue and chances to fix it after Nuclear Reset, Deep Freezing or Casting Spells are minimal. The device just has to be replaced. Routers fail for different reasons and have to be replaced like any other electronic device.

No, chasing fixes when none stick is the ultimate waste of time. A proper reset (regardless of what it is named) and waiting the suggested time limits works because of what you don't know, not because of what you do know.

If everyone simply dumped their router because of configuration errors (or, as I call them; inter-firmware interactions), there would be a lot of good routers in the landfills/RMA's today.

I also like how you ignore what RMerlin states too. Reset to test if it is the router or the network.

The Nuclear Reset resets everything in the network and is one of the best ways to test for a single piece of that puzzle (like the router, in our examples here).

Even with electrons not flowing, there are other devices that are still working that need the 'time out' to reset themselves. ISP modems/backends, etc.

Not to mention random and effectively untraceable static buildup in LAN cables, power cables, circuits, cases and any other connected surfaces that do need the 5/10/60 minutes to properly dissipate that static charge, if that is indeed the issue (the catch 22 is that you don't know if you don't try).

Your knowledge of component level router electronics seems limited to lab-based setups and theoretical applications?

For my customers, I have to deal with actual issues that are just as real too and not just state they need to buy new equipment, now.

Since you are so opposed to my working 'Nuclear Reset' process, I am waiting for your suggestions to accomplish the same thing 'initiated in a different way'.

If you can give a more direct way of getting our routers to a good/known state, I'm sure we'll all appreciate it.

 

[dosborne]

I see a trend here on SNB to recommend a reset for almost everything without even trying to help the user by asking the right questions.

There is a reason for that. The reason is it fixes a large majority of the issues. I know it did for me. Part of the issue is that configuration files are binary and are not portable between versions. Some systems create xml or txt files to store the configuration to make it cleaner when you load it bad into a previous or current firmware version. Unfortunately, with Asus, this is not the case and therefore a bit or byte out of place causes all sorts of strange behaviour. Why spend hours / days / weeks trying to help someone out with an issue that may not even be a real issue? If something is not working right on your router, step one should be to erase the configuration and start from scratch. It isn't particularly hard to create a text file or doc with screen captures of the setup you want. Yes, it may take you 20 minutes to reconfigure, but at least you know that your router is "clean" and can start troubleshooting and asking for help.

I too have many (35+) years experience as an electronic design, electrical engineer, programmer,... but I still needed to be told to reset and reconfigure manually, and it fixed my issues. That is why I have been trying to help others here lately, because I was helped.

Don't forget as well that many people use their router to perform tasks they were never designed to do. Running scripts, ad blockers, firewalls, media server, all sort of USB devices,.... These all potentially introduce memory issues (overuns, leaks, ...). I would tend to agree with you on some of the timings for a powered down state, but I'm not about to crap all over the people that work very hard either to provide better functionality or assistance to those in need. Many people don't know or understand the boot process of the router so the powerup wait times *are* important.

Do you need to do a full reset for EVERY update? No, that is up to you. Probably do it only if you encounter an issue. Should you do it before wasting everyone else's time? Yes.

 

[dosborne]

This so called Nuclear Reset is the ultimate waste of time.

Tell that to the people that thought they had turned their routers into bricks but now have working routers again.

 

[Pergola Fabio]

i am still on 384.9 , can i just update to this one?
dont wanne start from scratch...

 

[Val D.]

No, chasing fixes when none stick is the ultimate waste of time. A proper reset (regardless of what it is named) and waiting the suggested time limits works because of what you don't know, not because of what you do know.

Based on questions asked on SNB and issues mentioned I know for a fact that 99% of users on this forum have simple home network setup with one main router and some common devices connected to it wired or wireless. We are not talking about a corporate network of 1000 computers here. No corporate network uses consumer products as main hardware anyway. Router reset procedure is described in the User Manual and it takes 60 seconds to perform. No need to reinvent the wheel.

Merlin is absolutely correct to recommend a fresh start when installing his version of the firmware. This way he can filter out the firmware bugs from inherited issues from previous firmware installations and configurations. I don't see Merlin trying to reinvent the reset procedure though.

 

[dosborne]

i am still on 384.9 , can i just update to this one?
dont wanne start from scratch...

You need to determine if the need for the fixes, patches, additions in the new firmware outweigh the risk of having to manually reconfigure. I did a minor firmware update and had to. But, that was the first time in many years, and many firmware updates. I don't think anyone can say either way with 100% certainty if you will run into issues or not. If you don't want the risk, don't update.

 

[skeal]

You might try this in dnsmasq.add (from the Stubby "old" days)

server=/pool.ntp.org/1.1.1.1

Hey bud was that added in the dnsmasq.conf.add file or elsewhere?

 

[joe68000]

RT-AC5300 dirty flash over 384.11_2 - no issues at all.

Thanks RMerlin!

 

[skeal]

Known issue that's been there for months, Asus only recently fixed it on their end, but haven't released any new firmware since fixing it.



Works for me.

View attachment 18363

The main difference when enabling that option is a new firewall rule is created to redirect traffic. If you use any customized firewall rules, review your changes.

I was only using the top server text box, when I added the second server in the second server text box, it all works. Me and @Swistheater figured there was something odd happening after a firmware flash and the time not syncing at least once.

 

[scjr]

Based on questions asked on SNB and issues mentioned I know for a fact that 99% of users on this forum have simple home network setup with one main router and some common devices connected to it wired or wireless. We are not talking about a corporate network of 1000 computers here. No corporate network uses consumer products as main hardware anyway. Router reset procedure is described in the User Manual and it takes 60 seconds to perform. No need to reinvent the wheel.

Merlin is absolutely correct to recommend a fresh start when installing his version of the firmware. This way he can filter out the firmware bugs from inherited issues from previous firmware installations and configurations. I don't see Merlin trying to reinvent the reset procedure though.

But RMerlin has recommended on these forums, several times, a power cycle to complete an electrical reset. Just search ‘unplug’ and RMerlin as the author. So it’s obvious there are certain scenarios that this might help the end user. It can’t hurt.

https://www.snbforums.com/threads/spontaneous-reboot-should-i-be-worried.50139/#post-447486