Asuswrt-Merlin 380.58 is now available

[Marc Elser]

Hi Everyone,

I have 2 scripts in my /jffs/scripts directory called "firewall-start" (adds iptables rules) and "services-start" (starts nginx). Since several releases (also in 380.58) "services-start" which should start nginx is not executed anymore, only "firewall-start" (I can see firewall-start in the logs but not services-start). Also in the webui jffs scripts are set to enabled.

both scripts are set to be executable and also running /jffs/scripts/services-start from ssh shell works fine and starts nginx. Was there any change in script names or any other reason why services-start doesn't get executed? I also tried adding a sleep command at start of services-start to delay start of nginx but it really seems that it's simply not executed, so ading a sleep doesn't help at all. the script looks like this

#!/bin/sh

/opt/etc/init.d/S80nginx start

Any ideas?

 

[RMerlin]

Hi Everyone,

I have 2 scripts in my /jffs/scripts directory called "firewall-start" (adds iptables rules) and "services-start" (starts nginx). Since several releases (also in 380.58) "services-start" which should start nginx is not executed anymore, only "firewall-start" (I can see firewall-start in the logs but not services-start). Also in the webui jffs scripts are set to enabled.

both scripts are set to be executable and also running /jffs/scripts/services-start from ssh shell works fine and starts nginx. Was there any change in script names or any other reason why services-start doesn't get executed? I also tried adding a sleep command at start of services-start to delay start of nginx but it really seems that it's simply not executed, so ading a sleep doesn't help at all. the script looks like this

#!/bin/sh

/opt/etc/init.d/S80nginx start

Any ideas?

services-start might be executed before your USB disk is ready (and Entware available).

To confirm that the script does get executed, add a simple touch command before the nginx call to create an empty file in /tmp when it gets executed:

touch /tmp/000services-start

 

[malinois1]

Hi Folks,

for some reason my RT-AC87U keeps freezing from time to time. A few month ago I had the freezing issue with the stock ASUS firmware. This is why I came to ASUS Merlin in the first place. After having Merlin installed quite a while I tried again the latest Asus stock firmware. Still, freezing was an issue from time to time. Now, using the latest 380.85 Merlin release my router crashed three times in the last 7 days:

• first I experience lags in the internet connection (as I use very often Teamspeak it is easy to detect when the line goes down)
• then, my windows machine will show an exclamation mark on the network connection
• aaaaand the router is down: network dead, router unresponsive
• Syslog shows no sign of trouble --> it just stops logging (obviously as the network is dead)

Is there any fix for that? Is this related to the RT-AC87U? By the way: The RT-AC87 reboots every day at 4am - so its not like this machine is running for weeks without reboot.

The network crash happened most of the time during online gaming (Dirt Rally, DayZ, ...) -- don't know if it is related to the games.

It frustrates me so much this router...

From what I have gleaned from skimming over the forums is that the ac87U has a Quantenna's QSR1000 4x4 chipset which is known to be a problem. Many think Asus owes them a refund for selling them such a buggy router. If I was in your position I would trade the router in at amazon....should fetch about $100 ....and buy a known good router. I think the 88U is pretty well proven by this point. Google AC87u problems. Here is a link http://lmgtfy.com/?q=Asus+RT-AC87U+problems

 

[joegreat]

From what I have gleaned from skimming over the forums is that the ac87U has a Quantenna's QSR1000 4x4 chipset which is known to be a problem. Many think Asus owes them a refund for selling them such a buggy router. If I was in your position I would trade the router in at amazon....should fetch about $100 ....and buy a known good router. I think the 88U is pretty well proven by this point. Google AC87u problems. Here is a link http://lmgtfy.com/?q=Asus+RT-AC87U+problems

I completely disagree: In my case the AC87U runs like a charm from day one!
The reality is that some owners have problems, but the cause of the problems cannot be generalized. Most of the people without problem never show out in the postings - only the once with problems flood the forums with unspecific messages...

Now I updated to a tweaked version of Merlin's great 380.58 release - again without any issues.
By the way: I invested only 30 bucks (or "bugs") more then a AC86U (as my old one died) would cost and the update is worth much more then it costs!

 

[mgeoffriau]

By the way: I invested only 30 bugs more then I would for a AC86U (as my old one died) and the update is worth every cent!

Now that's a funny typo.

 

[malinois1]

I completely disagree: In my case the AC87U runs like a charm from day one!
The reality is that some owners have problems, but the cause of the problems cannot be generalized. Most of the people without problem never show out in the postings - only the once with problems flood the forums with unspecific messages...

Now I updated to a tweaked version of Merlin's great 380.58 release - again without any issues.
By the way: I invested only 30 bucks (or "bugs") more then a AC86U (as my old one died) would cost and the update is worth much more then it costs!

I don't disagree completely with you. After all it is how people are. People come to forums to complain more than praise. However it seems more people have problems with that chipset.

 

[Marc Elser]

services-start might be executed before your USB disk is ready (and Entware available).

To confirm that the script does get executed, add a simple touch command before the nginx call to create an empty file in /tmp when it gets executed:

touch /tmp/000services-start

Seems I was wrong and /jffs/services-start gets executed because the /tmp/000services-start got created. So it really seems that service-start gets executed as you said before entware. Is there any other startup file or a trick how to make nginx start after entware is available?

 

[joegreat]

Seems I was wrong and /jffs/services-start gets executed because the /tmp/000services-start got created. So it really seems that service-start gets executed as you said before entware. Is there any other startup file or a trick how to make nginx start after entware is available?

Hi,
The post-mount script is your friend here! See below my example script...

With kind regards
Joe

Spoiler: post-mount example script

#!/bin/sh
/usr/bin/logger -t START_$(basename $0) "started [$@]"
SCRLOG=/tmp/$(basename $0).log # creates a script specific log file in /tmp
touch $SCRLOG
NOW=$(date +"%Y-%m-%d %H:%M:%S")
echo $NOW "START_$(basename $0) started [$@]" >> $SCRLOG
#
if [ $1 = "/tmp/mnt/ac87u" ] # USB mounted - change it with your path
then
ln -sf $1/entware.arm /tmp/opt >> $SCRLOG
# Start Entware
/opt/etc/init.d/rc.unslung start >> $SCRLOG
#
# Add here what else you want to do after Entware started!
#
fi
#STOP Logging of script information
NOW=$(date +"%Y-%m-%d %H:%M:%S")
if [ "$?" -ne 0 ]
then
echo $NOW "Error in script execution! Script: $0" >> $SCRLOG
else
echo $NOW "Script execution OK. Script: $0" >> $SCRLOG
fi
/usr/bin/logger -t STOP_$(basename $0) "return code $?"
exit $?

 

[Matteo Guglielmi]

Seeing this as well: I happen to be uploading initial backup of my laptop to the cloud so heavy and sustained bandwith usage for several days now. Result? Router completely freezes every 24hrs or so. I have to recycle power.



Great to hear this is a known bug and already fixed by ASUS. Just have to wait for Merlin to incorporate it now. Thanks Matteo!

Issues are back with latest merlin firmware which is kind of unexpected since the firmware revision seems to be more recent than asus beta one.

 

[L&LD]

Issues are back with latest merlin firmware which is kind of unexpected since the firmware revision seems to be more recent than asus beta one.

That is impossible (the bolded text).

 

[Matteo Guglielmi]

That is impossible (the bolded text).

I based my assumption on the last 8 characters of the corresponding binary files:

merlin firmware: RT-AC87U_380.59_beta1-g3b6a743.trx

asus firmare: RT-AC87U_9.0.0.4_380_2695-g1b222cd.trx

but I can be wrong.

 

[RMerlin]

I based my assumption on the last 8 characters of the corresponding binary files:

merlin firmware: RT-AC87U_380.59_beta1-g3b6a743.trx

asus firmare: RT-AC87U_9.0.0.4_380_2695-g1b222cd.trx

but I can be wrong.

Those numbers are SHA hashes, and can't be compared. They don't mean anything beside reporting from which Git tag we compiled our respective firmwares - and they're compiled from different Git repositories, so they will never match.

 

[Matteo Guglielmi]

Those numbers are SHA hashes, and can't be compared. They don't mean anything beside reporting from which Git tag we compiled our respective firmwares - and they're compiled from different Git repositories, so they will never match.

So,

RT-AC87U_380.59_beta1-g3b6a743.trx is based on 3.0.0.4.380_2697 which I suppose is more recent than RT-AC87U_9.0.0.4_380_2695-g1b222cd.trx, am I correct?

 

[RMerlin]

So,

RT-AC87U_380.59_beta1-g3b6a743.trx is based on 3.0.0.4.380_2697 which I suppose is more recent than RT-AC87U_9.0.0.4_380_2695-g1b222cd.trx, am I correct?

The only difference between 2695 and 2697 was like 3 or 4 lines of code. Chances are they are only one or two days apart from one another, so from all intent and purposes, it's the same.

 

[OAM]

Hi,

I recently got a Linksys EA6900 and promptly changed the firmware from stock to the AsusWRT (Thank you Merlin!)

I started with v 380_58_0... everything worked great except three issues.

1> The 5GHz wireless would not connect when authentication was enabled. Without authentication, i was able to connect fine.
2> Router would start hanging up and dropping connection after couple of days of use. Reset, and all was fine for a couple of more days.
3> dlna and drive share was excruciatingly slow. I have a USB 2 HDD connected to the router. Its formatted as NTFS. I read on another post that adding a ext3 partition helps. Will try that. But am happy to get suggestions from others who have solved this issue.

Researching through the posts, I noticed number of similar issues reported with 58, hence I switched to 380_57... this resolved #2 above. Router was stable for a week+.. but #1 and #3 still persisted. I did the factory default reset etc. but no success.

So today i downgraded to 378_56_2... this has resolved #1. but alas #3 still persists.

I am very new to this Router and firmware and recognize this may just be due to novice mistakes on my part so would appreciate any help resolving these issues.

Thanks in advance.

 

[Matteo Guglielmi]

The only difference between 2695 and 2697 was like 3 or 4 lines of code. Chances are they are only one or two days apart from one another, so from all intent and purposes, it's the same.

Well, weirdly enough, 2697 has all the bugs (1 and 2 GB/sec network monitor spikes + router crashes) that 2695 does not have... In fact, I've just reinstalled the asus beta firmware and they are all gone.

Can this be the case?

I mean that 2697 did reintroduce back all the bugs?

 

[RMerlin]

I recently got a Linksys EA6900 and promptly changed the firmware from stock to the AsusWRT (Thank you Merlin!)

This hardware is not supported by the firmware.

 

[RMerlin]

Well, weirdly enough, 2697 has all the bugs (1 and 2 GB/sec network monitor spikes + router crashes) that 2695 does not have... In fact, I've just reinstalled the asus beta firmware and they are all gone.

Can this be the case?

I mean that 2697 did reintroduce back all the bugs?

No. As I said, there's almost no difference between 2695 and 2697. This is the sum of what's different between their respective /router/ folders:

diff -r asuswrt.380-2695-ac3200/release/src/router/extendno.conf asuswrt.380-2697-ac88/release/src/router/extendno.conf
1c1
< EXTENDNO=2695-g1b222cd
---
> EXTENDNO=2697-gb7c4567
diff -r asuswrt.380-2695-ac3200/release/src/router/Makefile asuswrt.380-2697-ac88/release/src/router/Makefile
3136a3137
> SRC_PRE_WLEXE = $(SRCBASE)/wl/exe
3167c3168
<     install $(SRC_WLEXE)/prebuilt/wl wlexe
---
>     install $(SRC_PRE_WLEXE)/prebuilt/wl wlexe
diff -r asuswrt.380-2695-ac3200/release/src/router/shared/defaults.c asuswrt.380-2697-ac88/release/src/router/shared/defaults.c
14,17d13
< #ifdef RTCONFIG_TCODE
< #include "tcode.h"
< #endif
<

They ain't even code changes, they're all related to the firmware build environment. Your issue most likely lies in the fact that Asus's traffic monitor is broken in many devices at the moment, and I have to disable that new codepath that they are using so it will at least report sensible values. Asus's firmware reports almost identical overblown values for both download and upload.

 

[Bamsefar]

I have a problem with my AC88 with 380.58 firmware: So far my router has hang twice in 49h, and when this occurs it is so bad that I have NO access and NO internet. Well I have nothing (so if this happens, no one can do nothing - well I guess if everything is to go wrong that is a good thing). The only way to solve this is a power cycle. And then it is just like before. After the first hang, I decided to reboot the router every night thru the firmware. That did not help, since I got another hangup about 49h later (it is actually very exact 49 hours - very strange).

So what have I been playing with? Well only my VPN server 1, so now I have closed that one (turned it off). I have been investigating Port Knocking for my VPN, but so far I have only collected information (no commands what so ever). So I am interesting in if closing my VPN server will "solve" this.

Why VPN concern then? Well My general thought is to route all my mobile traffic over my VPN connection when I am not home. I have this working, with a Androind Tasker task that turns on VPN client in my mobile when I leave my WiFi connection (that works just perfect by the way). So I end up with maybe 8-12h VPN connection active per day. Could this create this error when it hangs? I will also connect my wifes mobile like this, and most likly my parents router also and so on, if I can trust the VPN connection... (well if I end up with that much VPN usage I guess VPN port knocking is a bit unnecessary).

The bad thing with powercyle the router is that I can not get any information about what has happend. I have tried to connect with putty to my router when this happens, but that also fails (It does work under normal condition).

At some stage I could say: This looks like a DDOS attack, however direct after powercycle everything is back to normal. If it is a DDOS well then powercycle should not matter, right?

Oh by the way, I have a static IP address on internet so no DynDNS or anything (well my ISP uses DHCP for the static addresses also but that should not matter?).

 

[Matteo Guglielmi]

I have a problem with my AC88 with 380.58 firmware: So far my router has hang twice in 49h...

You are not alone.

Install the latest beta firmware 380.59_beta2 or the latest beta firmware from asus.