Asuswrt-Merlin 378.56 Beta 2 is out

[fryedchikin]

Merlin, I can provide a copy of my syslog if that is what you are looking for. I may have to edit out a few things such as ssh logins etc. It does show where ntp was not syncing until I had made the wan dns settings change that seemed to kick it off and then the time was corrected. Once fixed the problem does not come back. I only had it after flashing from 378.55 to 378.56beta1 and then when I flashed to 378.56beta2.

How can I get a copy to you?

I am able to recreate the issue by doing a full power cycle of the router. Will try to pull a syslog later.

 

[thelonelycoder]

I downgraded the firmware to 378.55 because of the ntp no sync . It got so bad that even a wan DNS change did not help. Now it runs fine again. I use the exact same scripts.

 

[fryedchikin]

I am able to recreate the issue by doing a full power cycle of the router. Will try to pull a syslog later.

I just captured a log for you. This is from a clean power cycle.

http://pastebin.com/T3UZZawa

 

[fafot]

I think I found a bug on the Administration-System page. I am still in 378.56bata 1 but I am not sure that this bug was fixed in beta 2:

At the bottom of the page when trying to delete the "client list" of "Specified IP addresses" - after deleting all of them and pressing Apply button I am receiving an error: "http://192.168.1.1 Fields cannot be blank". As a result once you specified up to 4 addresses to access the router, you can only delete 3 of them but at least one must be in the table! This is preventing access to the router from anywhere but the specified client!

 

[RMerlin]

I downgraded the firmware to 378.55 because of the ntp no sync . It got so bad that even a wan DNS change did not help. Now it runs fine again. I use the exact same scripts.

I can't fix bugs that I cannot reproduce, unless I get more info (so far I only received one syslog, and for that user the issue seem to be caused by a buggy dnsmasq customization).

Are you customizing dnsmasq by any chance?

 

[RMerlin]

I just captured a log for you. This is from a clean power cycle.

http://pastebin.com/T3UZZawa

You're the second user with the NTP issue to also have a problem with his dnsmasq configuration.

[
Jul 31 19:00:13 dnsmasq[420]: cannot open log /tmp/mnt/sda1/logs/dnsmasq.log: No such file or directory
Jul 31 19:00:13 dnsmasq[420]: FAILED to start up

Your dnsmasq fails to start at boot time because you are trying to access a USB disk that is not available yet. And since dnsmasq fails to start, then so does the NTP update.

Remove that dnsmasq customization and try again to reboot your router.

 

[fryedchikin]

Thanks RMerlin, I will try removing that customization as soon as I am able to reboot. However, that configuration was working fine with 378.55. I have the log location specified in dnsmasq.conf.add, is there a better method I should be using ?

 

[MXM]

RMerlin,

I have the RT-AC3200 running this Beta 2. My syslog is full of these messages:

getlargeresourcerecord: opt 65002 optlen 8 wrong

From what I was able to find, this problem was addressed back in 376.44 but it now seems to be back. Any chance you can take a look at it?

Thanks
M

 

[RMerlin]

Thanks RMerlin, I will try removing that customization as soon as I am able to reboot. However, that configuration was working fine with 378.55. I have the log location specified in dnsmasq.conf.add, is there a better method I should be using ?

It was working through luck. Probably ntp update managed to get done only after the USB disk got mounted (and dnsmasq succesfully started), so it didn't interfere with the NTP update process. Just a matter of timings.

Ideally, you'd want to use a postconf script instead of an add file. The postconf script should check if the USB disk exists. If it does, then append your logfile line to dnsmasq.conf. If not, run dnsmasq without it. That will make your dnsmasq run regardless of whether the USB disk is found ot not.

 

[RMerlin]

RMerlin,

I have the RT-AC3200 running this Beta 2. My syslog is full of these messages:

getlargeresourcerecord: opt 65002 optlen 8 wrong

From what I was able to find, this problem was addressed back in 376.44 but it now seems to be back. Any chance you can take a look at it?

Thanks
M

I can't find the "getlargeresourcerecord" string anywhere in the firmware source code. Are you sure it's not coming from something you installed through Entware?

 

[RMerlin]

Thanks RMerlin, I will try removing that customization as soon as I am able to reboot. However, that configuration was working fine with 378.55. I have the log location specified in dnsmasq.conf.add, is there a better method I should be using ?

Here is a postconf script that should work (just edit the path and filenames as appropriate):

#!/bin/sh
CONFIG=$1
source /usr/sbin/helper.sh

if [ -d "/mnt/sda1" ]; then
        pc_append "log-facility=/mnt/sda1/dnsmasq.log" $CONFIG
fi

 

[MXM]

I can't find the "getlargeresourcerecord" string anywhere in the firmware source code. Are you sure it's not coming from something you installed through Entware?

RMerlin,

I never used Entware at all.

Here's what the full log entry looks like...

Oct 21 21:22:02 mDNSNetMonitor: GetLargeResourceRecord: opt 65002 optlen 8 wrong

Thanks,
M

 

[MXM]

Maybe it can help:

http://answers.microsoft.com/en-us/...-8-wrong/c8ed8f34-7197-4c42-b9d3-aacae0073777

But this addresses Windows Event Log messages.

The messages I am seeing are in the router syslog...

Thanks.

 

[RMerlin]

RMerlin,

I never used Entware at all.

Here's what the full log entry looks like...

Oct 21 21:22:02 mDNSNetMonitor: GetLargeResourceRecord: opt 65002 optlen 8 wrong

This message should only be logged if you have Syslog set to Debug level. Check under Administration -> System what level you are using.

Also make sure you are running Beta 2, as previous releases were logging these with a lower priority level.

 

[thelonelycoder]

I can't fix bugs that I cannot reproduce, unless I get more info (so far I only received one syslog, and for that user the issue seem to be caused by a buggy dnsmasq customization).

Are you customizing dnsmasq by any chance?

I am using these entries:

# Google
    server=8.8.8.8
    server=8.8.4.4
# Logging
    log-facility=/tmp/mnt/sda1/logs/dnsmasq.log
    log-async
    log-queries
# AdBlocking
    address=/0.0.0.0/0.0.0.0
    ptr-record=0.0.0.0.in-addr.arpa,0.0.0.0
    addn-hosts=/tmp/mnt/sda1/hosts/hosts.blocked.txt
    addn-hosts=/tmp/mnt/sda1/hosts/blacklist.txt
# internal Servers
    address=/test/192.168.2.150
    address=/dev/192.168.2.160
    address=/aus/192.168.2.170
    address=/red/192.168.2.165
# ipset-dns redirections
    conf-file=/jffs/configs/dnsmasq.win10.conf

I am going to look into the dnsmasq.log errors at startup as soon as I have time to upgrade to beta 2 again. Probably on Sunday.

 

[RMerlin]

I am using these entries:

    log-facility=/tmp/mnt/sda1/logs/dnsmasq.log

Same problem as the two other users there. Look at the postconf script I posted as a safer way to use a USB-stored logfile.

The good news is that it solves the mystery of the randomly failed ntp update - everyone affected so far were storing their dnsmasq on a USB disk, which wasn't mounted at the time the initial dnsmasq was launched.

 

[fafot]

Merlin,
I am wondering if you can have a look at the Administration-System page bug that I wrote about in post # 130.
Tx

 

[HardCat]

Here is a postconf script that should work (just edit the path and filenames as appropriate):

#!/bin/sh
CONFIG=$1
source /usr/sbin/helper.sh

if [ -d "/mnt/sda1" ]; then
        pc_append "log-facility=/mnt/sda1/dnsmasq.log" $CONFIG
fi

I have made the changes as suggested and find that this method works much better! Thanks...

 

[MarioCaires]

RMerlin,

I never used Entware at all.

Here's what the full log entry looks like...

Oct 21 21:22:02 mDNSNetMonitor: GetLargeResourceRecord: opt 65002 optlen 8 wrong

Thanks,
M

I'm getting the same in 378.56 beta2 with a AC87U...

 

[RogueSpear]

Here is my dnsmasq.conf.add:

no-resolv
server=127.0.0.1#60053
server=127.0.0.1#60054

And here is my wan-start:

#!/bin/sh

logger -t $(basename $0) "started [$$]"

/bin/pidof dnscrypt-proxy > /dev/null 2>&1 || \
(

  # restart NTP client to eliminate 4-5 mins delay

  killall ntp
  sleep 5
  service restart_dnsmasq
  sleep 5
  service restart_ntpc
  sleep 5


  # wait up to 5 minutes to make sure the router has the correct time

  tmax=300
  i=0
  while [ $i -le $tmax ]
  do
  if [ "$(nvram get ntp_ready)" -eq "1" ]
  then
  break
  fi
  logger -t $(basename $0) "Waiting for correct time to be set."
  sleep 1
  i=`expr $i + 1`
  done


  # dnscrypt-proxy requires the correct time for certificate validation

  /jffs/bin/dnscrypt-proxy --local-address=127.0.0.1:60053 --ephemeral-keys --resolver-name=dnscrypt.org-fr --resolvers-list=/jffs/bin/dnscrypt-resolvers.csv --daemonize
  /jffs/bin/dnscrypt-proxy --local-address=127.0.0.1:60054 --ephemeral-keys --resolver-name=soltysiak --resolvers-list=/jffs/bin/dnscrypt-resolvers.csv --daemonize
)

The dnscrypt files are from lancethepants. This configuration worked without fail on .55. With beta 1 and 2 both, NTP will not work. It doesn't matter if I populate my hosts.add file with several pool.ntp.org entries, enter in an actual IP address that resolved to pool.ntp.org in the GUI, and it doesn't matter what DNS servers (my ISP, Google, Cisco, etc) that I enter into the GUI. I even tried turning on DNS filtering in AI Protection and then selecting the "No Filtering" option. I'm out of town and won't be home until this afternoon and I'm a bit nervous to be messing around remotely with this to get a syslog.. that I might bork my network and not have remote access. If this information doesn't help, I should be able to get a syslog later this evening.