OpenWrt Forum Archive

Topic: Update on Linksys WRT1900AC support

The content of this topic has been archived between 16 Sep 2014 and 7 May 2018. Unfortunately there are posts – most likely complete pages – missing.

Perhaps this may help those who have had wireless issues...

A few days ago I noticed wasn't able to connect to the 5gHz radio and flashing the trunk from a few days ago didn't work, nor the trunk from today, nor the RC2 build.  I then figured I'd go ahead and flash the Linksys firmware, then flash the trunk (and rebuild my pivot overlay partition)... which fixed the issue.

I don't know if users experiencing wireless disconnects have tried this, or if it will work to solve whatever the issue that's causing those disconnects is, but I thought it worth mentioning in case it does.

(Last edited by JW0914 on 27 Jun 2015, 21:20)

@kaloz

I sysupgrade -c trunk r46104 to r46133 today sad

RATS

It briked, no telnet or SSH And it did not keep my settings. I have no access to the OS. Can not install anything.

WAN came up, so I have the internet (thats all) It's briked.

jeremyjack wrote:

My CPU temperature hovers around 60 degrees, +/- 4 or so with the script running. I'm not sure if that's high or low compared to anyone else but it's working for me. If that seems high, I can adjust the CPU temperature checks to be a little more aggressive with the fan speeds.

That seems a bit high to me, as 60C is 140F... could someone verify what the normal temp range is?

deraol wrote:

Hello, sorry but I have just install the Chaos Calmer and can not telnet or even ping

The DEFAULT IP is 192.168.1.1 same?

Something is broken in trunk, it made my unit USELESS using sysupgrade today.

Gotta take it apart again sad

I swear, this router has a demon in it.

Wifi is up, so sysupgrade -c sorta worked in trunk but no dropbear. The unit is useless with out dropbear

(Last edited by gufus on 27 Jun 2015, 22:38)

jeremyjack wrote:
RickStep wrote:

Thanks for the quick turn around time for the code.

No problem at all! I was planning on doing something like this anyway at some point; a recent lockup led me to a new firmware version (and more lockups) and to your post.

The  flowchart was generated in response to [Chadster766]

Thanks, Chadster766 for getting the ball rolling on all of this smile

My previous record with RC2 was right around 48 hours before it locked up and needed a reboot, so I'm happy to report that I've been running it with this script for 7 days and 13 hours with some pretty heavy streaming and downloading. 7.5 days is going to be the limit right now because my cats broke into the server closet to play and managed to unplug a few things.

My CPU temperature hovers around 60 degrees, +/- 4 or so with the script running. I'm not sure if that's high or low compared to anyone else but it's working for me. If that seems high, I can adjust the CPU temperature checks to be a little more aggressive with the fan speeds.

I forgot to mention in my last post that I removed the following line from my crontab so that it wouldn't interfere with the new script:

 */5 * * * * /sbin/fan_ctrl.sh

For anyone else just now reading about this, I recently wrote a script to run all the time that effectively replaces /sbin/fan_ctrl.sh. It does checks of the system and adjusts fan speeds every 5 seconds and is based on this post:
https://forum.openwrt.org/viewtopic.php … 11#p280811

You can download it from here (save it to wherever you'd like):
https://raw.githubusercontent.com/jjack … control.sh

After downloading it, be sure to make it executable

chmod +x fancontrol.sh

To test the script:

./fancontrol.sh verbose

To just run the script in the background all the time:

./fancontrol.sh &

Please let me know how this works for any of you all.

Your script might have solved the lockup issue. Good going!

Just critical issue #20 left smile which is close to a month since discovery.

gufus wrote:
deraol wrote:

Hello, sorry but I have just install the Chaos Calmer and can not telnet or even ping

The DEFAULT IP is 192.168.1.1 same?

Something is broken in trunk, it made my unit USELESS using sysupgrade today.

Gotta take it apart again sad

And

gufus wrote:

I swear, this router has a demon in it.

Wifi is up, so sysupgrade -c sorta worked in trunk but no dropbear. The unit is useless with out dropbear

I experienced issues when I did both a sysupgrade, as well as a full flash (both saving and not saving current config), which resulted in issues... however, the fix is to flash the Linksys OEM firmware, then flash trunk from that (which worked for me).

(Last edited by JW0914 on 28 Jun 2015, 00:00)

Trunk is definitely not guaranteed to work on any given day. They are doing a lot of work now that CC is branched. Given the change to a new default C library and some security settings they're now enabling it's not a good idea to run trunk on a "production" device.

leitec wrote:

Trunk is definitely not guaranteed to work on any given day. They are doing a lot of work now that CC is branched. Given the change to a new default C library and some security settings they're now enabling it's not a good idea to run trunk on a "production" device.

Latest trunk runs fine... there's just a corruption issue that occurs if you flash it without flashing the OEM firmware first.  As far as what that corruption is, I have no clue.

(Last edited by JW0914 on 28 Jun 2015, 00:13)

Either way, now that there will eventually be a stable firmware for this router it's something to keep in mind.

leitec wrote:

Either way, now that there will eventually be a stable firmware for this router it's something to keep in mind.

I apologize, my intent wasn't to dismiss what you said, as it's an extremely important point.

No worries smile

leitec wrote:

Trunk is definitely not guaranteed to work on any given day. They are doing a lot of work now that CC is branched. Given the change to a new default C library and some security settings they're now enabling it's not a good idea to run trunk on a "production" device.

Ya, my fault.. I should know that.

leitec wrote:

Trunk is definitely not guaranteed to work on any given day. They are doing a lot of work now that CC is branched. Given the change to a new default C library and some security settings they're now enabling it's not a good idea to run trunk on a "production" device.

I think the majority of people are not reporting what they see as most don't run production and have no worries about doing a reboot. I have two friends that also run this router and we all experience problems of some sort. With all due respect to all the on-going work, this is the most unstable router we've ever run. At the very least a reboot is somewheres in the future. Trunk is seemingly more unstable as of late.

Hostname OpenWrt
ModelLinksys WRT1900AC
Firmware VersionOpenWrt Chaos Calmer r46133 / LuCI (git-15.176.44391-0ef5f3c) Kernel Version3.18.16
Local TimeSat Jun 27 20:59:21 2015
Uptime1d 2h 28m 32sLoad Average0.05, 0.03, 0.05

Seems ok here so far except no collectd. And running out of memory on large file transfers which is an ongoing problem that I wish I could get to the bottom of.

Been running some load tests today. What got me started on the tests, was CPU always seemed to range between 2% - 4% no matter what I was doing.

I even ran a test involving 5 torrents running at the same time with nearly a 1k peers, and processor was just pretty much sitting idle (maybe 2%) with a load of 0.1

My second load test was a little more harsh. The script below will run a loop, causing any CPU to instantaneously run at 100%

#!/bin/sh
for i in 1 2 3 4; do while : ; do : ; done & done

The script after about 2 minutes caused a load of 4.5 and processor utilization (according to top) of nearly 80%. My homegrown script showed utilization at 98%.

Anyhow, what did these tests prove? The router's hardware is completely over-kill for just about any home use. Here's the thing... Why in the heck would Liksys (Belkin), want to cripple the software so much? Heck, just add more flash, and it could be a gaming server... lol  An exaggeration, but I'm sure you get the point.

Anyhow, those are my thoughts.

(Last edited by davidc502 on 28 Jun 2015, 03:28)

about 2-3 months ago  @kaloz had mentioned that some trunk changes back then required the router to be flashed back to oem fw before flashing a newer openwrt revision. Though that requirement was only there for the next few builds backs then,,, could it be the same requirement has returned due to the recent changes in trunk code again?

(Last edited by alirz on 28 Jun 2015, 02:45)

alirz wrote:

about 2-3 months ago  @kaloz had mentioned that some trunk changes back then required the router to be flashed back to oem fw before flashing a newer openwrt revision. Though that requirement was only there for the next few builds backs then,,, could it be the same requirement has returned due to the recent changes in trunk code again?

It's recommended on the WRT1900ac wiki to flash back to OEM prior to flashing a new OpenWRT build.  I normally don't unless there's an issue that clearly isn't being resolved by reboot/reflashing OpenWRT.

For example, I lost the ability to connect to the 5gHz radio a few days ago and assumed a reboot would fix it... well several reboots and flashes later, the 5gHz radio was still showing a signal of -296 (which for all intents and purposes means it's off) and although it was clearly seen by both my NX6 and PC, I couldn't connect... leading me to flash OEM, then reflash Trunk today.

I've also encountered situations where writing OpenWRT via TTL won't allow the router to work the way it should (myriad of generic issues), thereby having to write the OEM fw to both the primary and backup image spaces.

If in doubt, flash OEM, then re-flash OpenWRT.

(Last edited by JW0914 on 28 Jun 2015, 03:01)

JW0914 wrote:
alirz wrote:

about 2-3 months ago  @kaloz had mentioned that some trunk changes back then required the router to be flashed back to oem fw before flashing a newer openwrt revision. Though that requirement was only there for the next few builds backs then,,, could it be the same requirement has returned due to the recent changes in trunk code again?

It's recommended on the WRT1900ac wiki to flash back to OEM prior to flashing a new OpenWRT build.  I normally don't unless there's an issue that clearly isn't being resolved by reboot/reflashing OpenWRT.

For example, I lost the ability to connect to the 5gHz radio a few days ago and assumed a reboot would fix it... well several reboots and flashes later, the 5gHz radio was still showing a signal of -296 (which for all intents and purposes means it's off) and although it was clearly seen by both my NX6 and PC, I couldn't connect... leading me to flash OEM, then reflash Trunk today.

I've also encountered situations where writing OpenWRT via TTL won't allow the router to work the way it should (myriad of generic issues), thereby having to write the OEM fw to both the primary and backup image spaces.

If in doubt, flash OEM, then re-flash OpenWRT.

I have been flashing quite a few different versions for about 3 months and the main rule at least for me is always flash factory first and make sure it is working before flashing anything else. There have been 2 different times that I have to do the triple boot and both times it worked. *Knocking on wood* I have a nice new usb>ttl cable still in it's bag...Just the way I like it.<G>
Not trying to be a smart a** just saying it has worked for me so far.

hi@experts :-)
i am looking for a solution:
i have running polipo on my wrt1900ac, but sometimes polipo crashes without errormessage and without removing it's pid and all my clients have no internet.
i need a loop something like that:
look with "ps | grep polipo"; if polipo is running, wait 3 min, if not: remove polipo.pid and start polipo
can somebody help me?
cu
p.

(Last edited by piphone on 28 Jun 2015, 09:40)

Has anyone configured dual/multi wan ports for wrt1900ac in openwrt and it worked?

I have tried several ways setting vlan but all tries are failed!

My OP system:

Firmware Version: OpenWrt Chaos Calmer r45885 / LuCI (git-15.151.52871-a835fc0)

Kernel Version:    3.18.14

_________________________________________________________________________________________

try: 1

vlan settings:

config switch
    option name 'switch0'
    option reset '1'
    option enable_vlan '1'

config switch_vlan
    option device 'switch0'
    option vlan '1'
    option vid '1'
    option ports '0 1 2 5'

config switch_vlan
    option device 'switch0'
    option vlan '2'
    option vid '2'
    option ports '3 6'

config switch_vlan
    option device 'switch0'
    option vlan '3'
    option vid '3'
    option ports '4 6'
   
result: only port 4 works.
        only port 4 can get ip from my ISP1 modem(192.168.1.x). there is no data recieved from port 3 (eth0.2) at the same time.
        but while setting port 4 unwired, port 3 can get ip from my ISP2 modem (192.168.8.x).
        There is no way to create an interface wan2 using eth1.x because there is no eth1.x in luci 'create interface' page. All vlan are eth0.x.
        It did not work if I added eth0.2 to wan2.

_______________________________________________________________________________________________

try: 2

vlan settings:

config switch
    option name 'switch0'
    option reset '1'
    option enable_vlan '1'

config switch_vlan
    option device 'switch0'
    option vlan '1'
    option vid '1'
    option ports '0 1 2 5'

config switch_vlan
    option device 'switch0'
    option vlan '2'
    option vid '2'
    option ports '4 6'

config switch_vlan
    option device 'switch0'
    option vlan '3'
    option vid '3'
    option ports '3 6'
   
result: only port 3 works.
        only port 3 can get ip from my ISP2 modem(192.168.8.x). there is no data recieved from port 4 (eth0.2) at the same time.
        but while setting port 3 unwired, port 4 can get ip from my ISP1 modem (192.168.1.x).
        There is no way to create an interface wan2 using eth1.x because there is no eth1.x in luci 'create interface' page. All vlan are eth0.x.
        It did not work if I added eth0.2 to wan2.

_______________________________________________________________________________________________

try: 3

vlan settings:

config switch
    option name 'switch0'
    option reset '1'
    option enable_vlan '1'

config switch_vlan
    option device 'switch0'
    option vlan '1'
    option vid '1'
    option ports '0 1 2 5'

config switch_vlan
    option device 'switch0'
    option vlan '2'
    option vid '2'
    option ports '3 4 6'

result:
port 4 is same as port 3. Either of them is wan port.
only port 4 or port 3 can get ip from my modem while one of them unwired. there is no way to create an interface wan2 using eth1.x because
there is no eth1.x in luci 'create interface' page. All vlan are eth0.x.

_______________________________________________________________________________________________

Layout:

Wrt1900ac is quite different from my previous router, e.g. tplink wdr4310, netgear wndr4300.

Those original wan ports are configured as eth0.2 while lan ports are eth0.1 and all of them are tagged to cpu .

but original wrt1900ac wan port(port 4) is configured as eth1 while lan ports are eth0.1 tagged to cpu but port 4 and 6 are also in eth0.2 not tagged to cpu.

It is very easy for me to create wan2 for tplink wdr4310 or netgeart wndr4300 and it always works.

____________________________________________________________________________________________


My conclusion:

There is no way to create a vlan using eth1 at the moment.

All vlan are in eth0. So it is impossible to create an eth1.x and dual/multi wan for wrt1900ac is currently impossbile...

Am I right?

(Last edited by coffeecat on 28 Jun 2015, 10:30)

northbound wrote:
JW0914 wrote:
alirz wrote:

about 2-3 months ago  @kaloz had mentioned that some trunk changes back then required the router to be flashed back to oem fw before flashing a newer openwrt revision. Though that requirement was only there for the next few builds backs then,,, could it be the same requirement has returned due to the recent changes in trunk code again?

It's recommended on the WRT1900ac wiki to flash back to OEM prior to flashing a new OpenWRT build.  I normally don't unless there's an issue that clearly isn't being resolved by reboot/reflashing OpenWRT.

For example, I lost the ability to connect to the 5gHz radio a few days ago and assumed a reboot would fix it... well several reboots and flashes later, the 5gHz radio was still showing a signal of -296 (which for all intents and purposes means it's off) and although it was clearly seen by both my NX6 and PC, I couldn't connect... leading me to flash OEM, then reflash Trunk today.

I've also encountered situations where writing OpenWRT via TTL won't allow the router to work the way it should (myriad of generic issues), thereby having to write the OEM fw to both the primary and backup image spaces.

If in doubt, flash OEM, then re-flash OpenWRT.

I have been flashing quite a few different versions for about 3 months and the main rule at least for me is always flash factory first and make sure it is working before flashing anything else. There have been 2 different times that I have to do the triple boot and both times it worked. *Knocking on wood* I have a nice new usb>ttl cable still in it's bag...Just the way I like it.<G>
Not trying to be a smart a** just saying it has worked for me so far.

Yup same here..no issues flashing various openwrt revision that I build my self every now and then. Just have to keep in mind that between openwrt builds you have to flash the sysupgrade and not the full image.  Full images are to go from oem fw to openwrt

coffeecat wrote:

Has anyone configured dual/multi wan ports for wrt1900ac in openwrt and it worked?

<snip>

My conclusion:

There is no way to create a vlan using eth1 at the moment.

All vlan are in eth0. So it is impossible to create an eth1.x and dual/multi wan for wrt1900ac is currently impossbile...

Am I right?

No, it should work. Your examples have two untagged VLANs going to the same CPU port (6). At least one needs to be tagged, and a corresponding eth1.X interface must be created. I don't know how to do it in LuCI, but the config file would look something like:

config interface 'wan3'
    option ifname 'eth1.3'
    option proto 'dhcp'

and so forth, assuming you tag VLAN 3 on port 6 but not VLAN 2. VLAN 2 would continue to appear untagged on eth1.

srd wrote:
leitec wrote:

Trunk is definitely not guaranteed to work on any given day. They are doing a lot of work now that CC is branched. Given the change to a new default C library and some security settings they're now enabling it's not a good idea to run trunk on a "production" device.

I think the majority of people are not reporting what they see as most don't run production and have no worries about doing a reboot. I have two friends that also run this router and we all experience problems of some sort. With all due respect to all the on-going work, this is the most unstable router we've ever run. At the very least a reboot is somewheres in the future. Trunk is seemingly more unstable as of late.

We can say that about the stock Linksys firmware, that likes to lock up, requiring a reboot, as well.

leitec wrote:

No, it should work. Your examples have two untagged VLANs going to the same CPU port (6). At least one needs to be tagged, and a corresponding eth1.X interface must be created. I don't know how to do it in LuCI, but the config file would look something like:

config interface 'wan3'
    option ifname 'eth1.3'
    option proto 'dhcp'

and so forth, assuming you tag VLAN 3 on port 6 but not VLAN 2. VLAN 2 would continue to appear untagged on eth1.

It works!  Thank you very much!

It cannot be done in LuCI because the eth1.3 shows as eth0.3 inLuCI, but it can be set under console.

Here is my /etc/config/network file (Dual wan settings for wrt1900ac openwrt):

___________________________________________________________________________________________

config interface 'loopback'
    option ifname 'lo'
    option proto 'static'
    option ipaddr '127.0.0.1'
    option netmask '255.0.0.0'

config globals 'globals'
    option ula_prefix 'fd79:09a9:45d8::/48'

config interface 'lan'
    option ifname 'eth0'
    option force_link '1'
    option type 'bridge'
    option proto 'static'
    option netmask '255.255.255.0'
    option ip6assign '60'
    option ipaddr '192.168.1.1'

config interface 'wan'
    option _orig_ifname 'eth1'
    option _orig_bridge 'false'
    option proto 'dhcp'
    option ifname 'eth1'

config interface 'wan6'
    option ifname 'eth1'
    option proto 'dhcpv6'

config switch
    option name 'switch0'
    option reset '1'
    option enable_vlan '1'

config switch_vlan
    option device 'switch0'
    option vlan '1'
    option vid '1'
    option ports '0 1 2 5'

config switch_vlan
    option device 'switch0'
    option vlan '2'
    option vid '2'
    option ports '4 6'

config switch_vlan
    option device 'switch0'
    option vlan '3'
    option vid '3'
    option ports '3 6t'

config interface 'wan2'
    option ifname 'eth1.3'
    option proto 'dhcp'

________________________________________________________________________________________

WAN

eth1
Uptime: 0h 9m 53s
MAC-Address: 00:00:00:00:00:00
RX: 96.56 MB (74235 Pkts.)
TX: 4.90 MB (43022 Pkts.)
IPv4: 192.168.1.2/24
     
WAN2

eth1.3
Uptime: 0h 9m 58s
MAC-Address: 00:00:00:00:00:00
RX: 1.68 KB (17 Pkts.)
TX: 8.81 KB (96 Pkts.)
IPv4: 192.168.8.138/24

________________________________________________________________________________________

Thank you again!

The rest job is to set iptable for routing strategy...

(Last edited by coffeecat on 29 Jun 2015, 01:29)