New Kong Mod Release - r18010 [Regular & King]

85 posts / 0 new
Last post
M4rv1n
M4rv1n's picture
mac913 said: On DD-WRT site

mac913 said: On DD-WRT site builds 18000, 18007 & 18024 K2.6 has broken SSH. I didn't test SSH on the vpn-build since I ran into QoS issues and currently trying to get seperate VLAN working on build 17865, no luck so far. Does SSH work with vpn-kong build 18010?

I try kingkong build and dlna/samba build 18010 version and for me ssh works correctly.

mac913
mac913's picture
Thanks guys for the feedback

Thanks guys for the feedback on SSH. When I get sometime I will load 18010 and test.

Kong.... Thanks for all you work!!

ReDaLeRt
ReDaLeRt's picture
 

 

MadRocker said: @ReDa there is no way in hell I can get dd-wrt.v24-15508_NEWD_mini_asus.trx to work or even get access to my RT-N16, have to unbrick it each time, how do you get this to work on yours ? ...00MAS R01 EL have see it as rev 2 will check my model version again after I unbrick it.

 

Avoid at any cost "mini" versions on 802.11n routers. I'm afraid that those early builds do not have the wifi N driver fully working.

I use the "Big" version without any issues. Just reset to defaults nvram before submitting the firmware file through the WebGUI. Then, after it reboots, reset another time the nvram.

 

I have nvram backups with daily-use setups for every single build that had test on my RT-N16. Just submit the firmware and the restore nvram. In less than 7 minutes I have the router running fully configured, in case of downgrading or re-testing a build.

There is an issue with nvram backups/restores about the SoC clocks. I have every time of changing the firmware and restore nvram to setup the underclocking (440,220,110)...

 

Best regards,

ReDaLeRt
ReDaLeRt's picture
M4rv1n said: Hi, I have

M4rv1n said: Hi, I have installed this build on RTN16 and I have problem with port forwarding on loopback. Port forwarding on client of the network works, but if I forward a port to transmission on the router, transmission say that the port is closed. I have tried also to add to startup firewall script iptables -t nat -I POSTROUTING -o br0 -s 192.168.1.0/24 -d 192.168.1.0/24 -j MASQUERADE but nothing change. Any idea?

Hi,

Put in "Firewall commands", on WebGUI:

iptables -I INPUT 1 -p tcp --dport 6889 -j logaccept

Change the port at your taste.

 

Best regards,

zoomlink
zoomlink's picture
MadRocker said: @Zoom any

MadRocker said: @Zoom any idea if the power dipped or spiked during this time, yup it is strange.

Mad..

The first reboot at 6:00ish AM, I have no evidence of a power outage/spike/dip, I know there was no outage because I did not find any of my appliance clocks blinking.

The second time that I happened I was awake and I did not notice anything either.

I should mention that I have a 1500VA UPS/Battery Backup/Voltage "Cleaner" as the power source for all my router/switch/cablemodem/smart-home stuff.  So typically these devices do not get affected by power events unless these are very prolonged.

ReDaLeRt
ReDaLeRt's picture
@Kong:

@Kong:

Can you upgrade the build to include this change, please? http://svn.dd-wrt.com/changeset/18028

Thank you in advance.

ReDaLeRt
ReDaLeRt's picture
Kong said: Hi ReDaLeRt, just

Kong said: Hi ReDaLeRt, just run: mkdir -p /jffs/lib/opkg I normally do this when automounting, so I guess you don't have a drive with jffs attached. I also flashed (via network cable) from r17940 with FF8 and didn't have a problem.

I guess it could be impossible to run OPKG on that location:

# mkdir -p /jffs/lib/opkg
cannot create directory `/jffs/lib': Read-only file system

 

There is any way you can compile this feature within KingKong builds?

http://www.dd-wrt.com/phpBB2/viewtopic.php?p=657995#657995

Thank you.

M4rv1n
M4rv1n's picture
ReDaLeRt said:

ReDaLeRt said:

M4rv1n said: Hi, I have installed this build on RTN16 and I have problem with port forwarding on loopback. Port forwarding on client of the network works, but if I forward a port to transmission on the router, transmission say that the port is closed. I have tried also to add to startup firewall script iptables -t nat -I POSTROUTING -o br0 -s 192.168.1.0/24 -d 192.168.1.0/24 -j MASQUERADE but nothing change. Any idea?

Hi, Put in "Firewall commands", on WebGUI: iptables -I INPUT 1 -p tcp --dport 6889 -j logaccept Change the port at your taste. Best regards,

Thank you now the port is open

M4rv1n
M4rv1n's picture
wabe said:

wabe said:

M4rv1n said: I have tried all build of all version, but port forwarding to the router doesn't work! Transmission & amuled still have port closed.

Check out the two firewall rules created in the script S98pound (pound.wanup). These work fine on this build but obviously you need to adjust the port numbers for transmission. And SSH works fine for me (kingkong build on ASUS RT-N16)

Thank you for your reply. Where I can find that script?

zoomlink
zoomlink's picture
@Kong-Santa

@Kong-Santa

Any chance of you pumping out a 18047 build and putting it under the tree for us?

:)

If not... enjoy yourself.

ReDaLeRt
ReDaLeRt's picture
@Kong-Santa

@Kong-Santa

+1. ^^

Merry Christmas.

Teknowiz
Teknowiz's picture
Not sure why some of you are

Not sure why some of you are having wireless performance issues with the latest King Kong build but for me the latest build with updated drivers resolved the performance issues. I was seeing before my Asus RT-N16 would drastically drop wireless speeds after 24 hours but with the new build, speeds are still great after 7 days uptime.

Thanks for your hard work Kong and have a Merry Christmas.

ReDaLeRt
ReDaLeRt's picture
Hi,

Hi,

I'm studying this issue using a spectrum analyser (Ubiquity Airview-EXT) and some improvements on router's environment.

I'm afraid that is a major problem than just a bug on wifi driver. The results are pretty nice since the performance get a jump up with some minor changes on the antennas and its position to the objects surrounding.

I get the "wl interference 2" off because, after some analysis on 2.4GHz band around my house during several days, is almost useless and could reduce performance based on algorithms working over false statements (duty cycle of neighbours WLAN are very low and almost within the noise floor). Even my microwave spreads a wide burst of interference (around channel 10) far away from the channel selected on the router within this new benchmark ("5"). No Bluetooth devices present and intensive to get caught frequently on scans maiden so far.

Besides that, I can't yet understand why the noise floor is always near -81 dB on the router. There is no evidence on spectral analysis (medium amplitude of noise floor is near -95 dB). The only option left is the TX/RX error/retrieving levels are too high to achieve good radio performance.

Maybe an alternative to this theory is the sometimes low RSSI and wide amplitude of the stream received on the client. I say this because it could be a faulty method to determine the best combination of antennas and its polarization to achieve best performance over the 2x2 MIMO method on the radio embedded. This could generate the symptom of low speed link reported on the router over low traffic between it and the client. With some huge packet flow flying between them the link speed improve, possibly by some change on the streams arrangement.

The problem may not be present on the driver but in the radio.

Is my case similar to the ones having also reduced wifi performance?

Best regards,

zoomlink
zoomlink's picture
@Red

@Red

Thanks for the in-depth testing. I continue to chase the cause of my "unscheduled" reboots every 24 hours or so.

After a reboot this morning around 10:00am, I went ahead and reversed my overclock of the RT-N16 from 532 to 480. I have also, turned off WMM (based on feedback from another forum). I have turned-off interference.

So let's see how it goes for the next 24 hours.

zoomlink
zoomlink's picture
Report from loading r18010

Report from loading r18010 KingKong on a brand new RT-N16v2

USB HDD - 2.5TB WD 3.5" partitioned as follows:
p1=Label: optware (1024 MB)
p2=Label: swap ( 256 MB)
p3=Label: jffs (1024 MB)
p4=Label: data ( 2.2 GB)

Loaded eko_17084_mini.trx
teleted into router and did 'erase nvram' + 'reboot'
Loaded KingKong_18010.bin (with reset option selected)
Once up and running, went into the router, entered basic configs for router and wireless. Hit save + apply.

When it settled down, I went ahead and plugged in the USB drive.

It mounted the three non swap partitions to: /tmp/mnt/disc1-part2, 3 & 4

I logged in to the router using SSH.

Did a 'blkid' and got the UUIDs (all letters appeared in lower case)

Went to the GUI, entered the appropriate UUIDs in the corresponding mount points for /opt, /jffs and /mnt. Hit Save + Apply.

NO DICE... it does not like the lower case letters. So I manually changed all the lower case letters to CAPs. Hit Save + Apply.

I rebooted again.

Now the auto-mount worked. BUT, when I logged in via SSH and went to /mnt to create directories, it would no let me create directories.

So I just got pissed and from / I did a chmod 777 /mnt - That did it, now I am able to create directories. Don't know exactly why this happened.

At this point I enabled ProFTD, SAMBA and MiniDLNA.

Saved + Applied and Rebooted.

Finally everything came up as expected.

Wireless was configured as AP, N-Only, Auto Channel, 40Mhz upper, Security: WPA2 Personal Mixed/AES. Disabled WMM, left everything else on defaults.

Router reported 300 initially, when I connected from my Laptop with an Atheros NIC, it went down to 270 but it was solid at 270 for about 4 hours.

After about 1 hour went head and ran the OTRW install script (prep_optware).

After the install, I disabled the following:

S10swap
S35automount
S55portmap
S56unfsd
S80samba

Rebooted AGAIN.

Now all seems to be working as expected, including Optware security/anti-hacking enhancements (Asiablock, Birmablock, Stophammer and Stophack).

Just wanted to post the play by play with this build on a brand new router out of the box.

So far no issues to report regarding my prior pains on the older v1 router.

ReDaLeRt
ReDaLeRt's picture
@zoomlink

@zoomlink

Try underclocking the router for 440MHz on the SoC. (440,220,110)

Best regards.

Kong
Kong's picture
@zoomlink, your crashes might

@zoomlink, your crashes might come from a bad power supply.

@all, I have monitored signals for the past 2 weeks and while I didn't have dropping issues, the signal is still not as good as in 15000 releases.

In addition to that my 5100AGN adapter is not able to do 300Mbps mode with the latest wireless drivers although signal strength is not the issue.

So right now I'm trying to build a new image which includes drivers from 15000.

Teknowiz
Teknowiz's picture
@ Kong

@ Kong

I tested using 5100AGN on my laptop and I was getting 300 MBps on and off averaging around 270.

zoomlink
zoomlink's picture
@Kong

@Kong

I my new router intall, I have some ask/suggestions for the SERVICES>NAS page:

If I am running OTRW and decide to install and run Samba35 + Samba35-Swat, AND
I am running MiniDLNA from the router's GUI, then every time I do a change to the MiniDLNA configuration and apply it, this action causes my OTRW Samba to get killed.

In other words, I cannot choose to run Samba from an Optware Script and runt MiniDLNA from the GUI or the other way around because it is an ALL OR NOTHING thing.

I basically had to create a script called S99minidlna with the following code:
==============================
#!/bin/sh
# /opt/etc/init.d/S99minidlna
# MiniDLNA start-up script
NAME=minidlna
NAME="$0"

if [ -n "`pidof minidlna`" ] ; then
optlog "S99minidlna: Stopping MiniDLNA."
killall minidlna
sleep 2
optlog "S99minidlna: Re-starting MiniDLNA."
/usr/sbin/minidlna -R -f /opt/etc/minidlna/minidlna.conf
else
optlog "S99minidlna: MiniDLNA is not running, Starting it now."
/usr/sbin/minidlna -R -f /opt/etc/minidlna/minidlna.conf
fi
===============================

Is there a way to enable compare the enable/disable state when the GUI page is loaded and if there was no change on either MiniDLNA section or the SAMBA section, then don't do anything... don't kill the process if nothing was changed at the GUI. I recall a similar behavior with the Privoxy settings and you fixed it somehow.

zoomlink
zoomlink's picture
Ok guys... more issues with

Ok guys... more issues with the new install. Now with the USB drive.

Question: Are there any known issues related to handling partitions greater 2TB?

My /mnt partition (P4) is giving me major headaches I formatted it as an EXT2 but is acting up only allowing me to create ONE directory and if I try to create a new one it gives me an I/O error.

This is output from 'dmesg':

attempt to access beyond end of device
sda4: rw=0, want=3727281248, limit=584087552
attempt to access beyond end of device
sda4: rw=0, want=3727281248, limit=584087552
attempt to access beyond end of device
sda4: rw=0, want=3727281248, limit=584087552
CPU 0 Unable to handle kernel paging request at virtual address 00000004, epc == 800e5544, ra == 800e8f70
Oops[#1]:
Cpu 0
$ 0 : 00000000 1100bc00 00000000 00001000
$ 4 : 87f92ee0 87f92ee0 00000000 00000009
$ 8 : 0000000c 80364ae4 87f92160 00001000
$12 : 810eeee0 838c8cc0 00000000 00000000
$16 : 87f92ee0 87f92ee0 87f92ee0 00000001
$20 : 00000001 810eeee0 85104d34 0000000c
$24 : 00000000 fffffffc
$28 : 87f8e000 87f8fbe8 0000000e 800e8f70
Hi : 00000000
Lo : 00000000
epc : 800e5544 Not tainted
ra : 800e8f70 Status: 1100bc03 KERNEL EXL IE
Cause : 00000008
BadVA : 00000004
PrId : 00019740
Modules linked in: nf_nat_pptp nf_conntrack_pptp nf_nat_proto_gre nf_conntrack_proto_gre snd_seq_oss snd_seq_midi snd_seq_midi_event snd_seq snd_pcm_oss snd_mixer_oss snd_usb_audio snd_hwdep snd_usb_lib snd_pcm snd_timer snd_rawmidi snd_page_alloc snd_seq_device snd soundcore ext2 mbcache etherip usb_storage sd_mod scsi_wait_scan scsi_mod ohci_hcd uhci_hcd ehci_hcd usbcore switch_robo switch_core bcm57xx
Process pdflush (pid: 79, threadinfo=87f8e000, task=87fc39a8)
Stack : 00000001 87f92ee0 00011210 8113ed80 800583f8 805199c0 00000000 85104dcc
855e86c8 00f4d958 87f92160 87f5d02c 87b70a2c 87f92ee0 87f92ee0 800e8f70
00000001 87f92ee0 00000000 00000000 800acfd4 87f92160 87b6bc00 87f92fa0
87f5d02c 81084560 87b70a2c 87f8fce0 00000001 00000001 00000001 810eeee0
87b70a2c 87f92ee0 00000001 00000001 00000001 810eeee0 800a4ec0 800a5a7c
...
Call Trace:[][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][]

Code: ae000004 8e020008 8e03001c 00039242 8c430044 8c420040 000325c0 00021242
nf_conntrack: table full, dropping packet.
nf_conntrack: table full, dropping packet.
nf_conntrack: table full, dropping packet.
nf_conntrack: table full, dropping packet.
attempt to access beyond end of device
sda4: rw=0, want=3780776384, limit=584087552
attempt to access beyond end of device
sda4: rw=0, want=2354253064, limit=584087552
attempt to access beyond end of device
sda4: rw=0, want=683749640, limit=584087552
attempt to access beyond end of device
sda4: rw=0, want=3149421472, limit=584087552

Kong
Kong's picture
@zoomlink, yes currently

@zoomlink, yes currently drives >2TB are not supported.

I just tested my 300Mbps issue again, seems like a reset did it.

zoomlink
zoomlink's picture
@Kong,

@Kong,

Is there a command I can give to shrink the size of that partition to just under 2TB? The device is now deployed remotely and I will not be able to reach it physically until tomorrow.

If not, I will have to wait until I can connect the device and use Gparted to change the size of the partition.

zoomlink
zoomlink's picture
One more thing... can we use

One more thing... can we use EXT4? is it a good idea?

ReDaLeRt
ReDaLeRt's picture
Hi

Hi

@Kong:

Can you compile 2 build, one of which with the new driver?

After disabled interference mitigation I'm now getting improved wifi speed.

Best regards,

Axel
Axel's picture
@kong. I am running 17670

@kong. I am running 17670:17737M on an Asus RT-N16. There's a weird NAT issue in the following double NAT configuration. Perform factory reset using the webui. Set WAN adapter to DHCP. It receives and address on the 10.200.0.0/24 network. LAN adapter is 10.86.41.65 on the 10.86.41.64/26 network. LAN DHCP server enabled. Internet gateway is 10.200.0.254 (Linksys WRT310n DD-WRT connected to a cablemodem).

My laptop is behind the Asus router. I get DHCP address fine. I can ping the Asus router. I can nslookup fine. I cannot ping any address on the 10.200.0.0/24 network or on the internet (8.8.8.8 times out). Web surfing does not work either.

If I telnet into the Asus router, everything works as expected from inside the router.

Now here's the really weird part. If I change the LAN segment from 10.86.41.64/26 to 192.168.1.0/24 and set the Asus router address to 192.168.1.1, my laptop's internet access is restored.

Why on earth would the NAT fail if I used two class A subnets and work if I use a class A and a class C?

I am about to put 100 routers in production using this firmware but this NAT issue is a showstopper for me.

I would also like to know if I should use 17670:17737M or your latest 18000+ revision? I am not using the wifi adapter for this application, just using the routers as internet gateways and OpenVPN gateways. In the future I might want to use 3G on USB for WAN backup and optware on a thumbdrive for several other uses.

Thanks in advance.

zoomlink
zoomlink's picture
Ok guys...

Ok guys...

I was a little concerned with the 'nf_conntrack: table full, dropping packet.' messages on my dmesg file.

After a little searching I found this post which seems to explain an issue with iptables.

http://blog.sam-pointer.com/tag/nf_conntrack:tablefull,droppingpacket

@Kong: Is there anything we can put in the system to avoid the what is explained in the the link above?

PetervdM
PetervdM's picture
@zoomlink:

@zoomlink:
open a terminal session and type

nvram set ip_conntrack_max=xxxx
nvram commit
sync;reboot

replace xxxx with 256

zoomlink
zoomlink's picture
@Peter

@Peter

Thanks. I am suffering because I am remote to this darn router and will not be able to touch it until later tonight.

Isn't this a symtom of a DoS attack?

Finaly, what command can I give this damn box to FORCE it to reboot.

I have done /sbin/reboot and it ignores it.

PetervdM
PetervdM's picture
@zoomlink

@zoomlink
just sync;reboot on the commandline. or you can use the web interface, tab administration, tab management, button reboot router at the bottom of the page.

ReDaLeRt
ReDaLeRt's picture
 

 

PetervdM said: @zoomlink: open a terminal session and type nvram set ip_conntrack_max=xxxx nvram commit sync;reboot replace xxxx with 256 <= xxxx <= 65535, i've set it to 16384

 

Mine (18010 KingKong):

# nvram show | grep  ip_co
size: 30367 bytes (2401 left)
ip_conntrack_tcp_timeouts=3600
ip_conntrack_max=4096
ip_conntrack_udp_timeouts=120

 

Running about 50 torrents on Transmission (locally) and heavy Web traffic (clients). No issues. Also, I'd never get over 1600 simultaneous connections.

 

zoomlink said: @Peter Thanks. I am suffering because I am remote to this darn router and will not be able to touch it until later tonight. Isn't this a symtom of a DoS attack? Finaly, what command can I give this damn box to FORCE it to reboot. I have done /sbin/reboot and it ignores it.

 

Just type "reboot" (without " ").

 

Best regards,

Pages