What's new

Custom firmware build for R7800 v. 1.0.2.22SF

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Did you do a factory settings?
After that restore the settings for best result.

In general, not so obligatory. I recommed this procedure mainly for R7500 in my readme (too different drivers with stock firmware).

R7800: should be compatible with stock, so not absolutely necessary for R7800.

Voxel.
 
In general, not so obligatory. I recommed this procedure mainly for R7500 in my readme (too different drivers with stock firmware).

R7800: should be compatible with stock, so not absolutely necessary for R7800.

Voxel.
I installed 22SF and I have a few issues.
When I do a reboot my 2.4 is not coming up. I have to start it from the gui.
I also did a speed test at dslreports and I am getting an F for the buffer bloat. Quality reads A+. Ping time is really good though!
 
I installed 22SF and I have a few issues.
When I do a reboot my 2.4 is not coming up. I have to start it from the gui.
I also did a speed test at dslreports and I am getting an F for the buffer bloat. Quality reads A+. Ping time is really good though!

Thats why i say factory resett.
 
I installed 22SF and I have a few issues.
When I do a reboot my 2.4 is not coming up. I have to start it from the gui.
I also did a speed test at dslreports and I am getting an F for the buffer bloat. Quality reads A+. Ping time is really good though!
Well, Buffer bloat: Specific stuff. With the same dslreports I am getting:

Overall: A+
BufferBloat: A
Quality: A+

The same router, the same firmware, different Internet line.

I had one feedback that man, he had BufferBloat “C” with stock FW and has got BufferBloat “B” with my build.

I’d suggest you to do some fine tuning. Check for example this hint:

https://www.bufferbloat.net/projects/cerowrt/wiki/Enable_ECN/#enabling-ecn-on-linux

i.e.
1. Enter to your router console (by SSH or if you did not set it yet by telnet, enabling telnet from http://routerlogin.net/debug.htm).
2. Run the command from console:

Code:
sysctl -w net.ipv4.tcp_ecn=1

3. Test by dslreports again.

If it will be better, you can correct the file /etc/sysctl.conf changing this value.

P.S. Sometimes it is really better to do reset ;-) Maybe.

Voxel.
 
Well, Buffer bloat: Specific stuff. With the same dslreports I am getting:

Overall: A+
BufferBloat: A
Quality: A+

The same router, the same firmware, different Internet line.

I had one feedback that man, he had BufferBloat “C” with stock FW and has got BufferBloat “B” with my build.

I’d suggest you to do some fine tuning. Check for example this hint:

https://www.bufferbloat.net/projects/cerowrt/wiki/Enable_ECN/#enabling-ecn-on-linux

i.e.
1. Enter to your router console (by SSH or if you did not set it yet by telnet, enabling telnet from http://routerlogin.net/debug.htm).
2. Run the command from console:

Code:
sysctl -w net.ipv4.tcp_ecn=1

3. Test by dslreports again.

If it will be better, you can correct the file /etc/sysctl.conf changing this value.

P.S. Sometimes it is really better to do reset ;-) Maybe.

Voxel.
That didn't seem to work. I still got an F on the bloat test.

I did a reset last night to try to fix the 2.4 problem which seems to have worked.
 
From an A to a C.

If you can, try to comment by "#" symbols the lines with "echo" commands in init script /etc/init.d/optware:

Code:
. . .
start() {
    # Socket buffers
#    echo 1048576 > /proc/sys/net/core/rmem_max
#    echo 1048576 > /proc/sys/net/core/wmem_max
. . .

and reboot router after this. Then again bloat test after reboot.

Voxel.
 
If you can, try to comment by "#" symbols the lines with "echo" commands in init script /etc/init.d/optware:

Code:
. . .
start() {
    # Socket buffers
#    echo 1048576 > /proc/sys/net/core/rmem_max
#    echo 1048576 > /proc/sys/net/core/wmem_max
. . .

and reboot router after this. Then again bloat test after reboot.

Voxel.
How do I edit the file in telnet? I don't know the command to edit in Linux
 
How do I edit the file in telnet? I don't know the command to edit in Linux
Well... "vi" is the best editir ;-) I guess you don't know it...

So the way is: copy this file to USB stick, then edit the file from stick (attached to computer). Then copy it from the stick to router. Something like:

Code:
cp /etc/init.d/optware /mnt/sda1/

then edit is attaching to computer, then attach it back to router and:

Code:
cp -f /mnt/sda1/optware /etc/init.d/optware

Voxel
 
Well... "vi" is the best editir ;-) I guess you don't know it...

So the way is: copy this file to USB stick, then edit the file from stick (attached to computer). Then copy it from the stick to router. Something like:

Code:
cp /etc/init.d/optware /mnt/sda1/

then edit is attaching to computer, then attach it back to router and:

Code:
cp -f /mnt/sda1/optware /etc/init.d/optware

Voxel
I don't have Linux for my computer.
I guess I'll have to d/l a distro and boot from the cd.
 
If you can, try to comment by "#" symbols the lines with "echo" commands in init script /etc/init.d/optware:

Code:
. . .
start() {
    # Socket buffers
#    echo 1048576 > /proc/sys/net/core/rmem_max
#    echo 1048576 > /proc/sys/net/core/wmem_max
. . .

and reboot router after this. Then again bloat test after reboot.

Voxel.
Hi Voxel!
This editing, anything not affect for me.
and this
Code:
sysctl -w net.ipv4.tcp_ecn=1
nothing no change. My tests (2,4 GHz) randomly jump from A+ A A+ to B C B and etc, that I would not change. :(

P.S. But speed of my WIFi connection, I am fully satisfied. And of course, also speed of your firmware .:D
 
Bufferbloat and its effects can happen anywhere along the route between you and your remote site (in this case, the dslreports test site). The router isn't the only location. Cablemodems for instance will have their own buffers there.

The best way to get good result along most of the route (provided the ISP isn't overloaded) is to cap your bandwidth a few percents below what your ISP is giving you, ensuring that you will never fill up the buffers in your modem/router/PC.
 
Bufferbloat and its effects can happen anywhere along the route between you and your remote site (in this case, the dslreports test site). The router isn't the only location. Cablemodems for instance will have their own buffers there.

The best way to get good result along most of the route (provided the ISP isn't overloaded) is to cap your bandwidth a few percents below what your ISP is giving you, ensuring that you will never fill up the buffers in your modem/router/PC.
That sounds good. On the R7800, it runs an Ooakla speed test by default. So I'll try to set the bandwidth manually before I go through all the editing etc. I just haven't used Linux since 2000. So I forget all that stuff. I also have a very active network ie. Xbox, PS3, computers and if I mess it up I catch a lot of shxx!
 
Bufferbloat and its effects can happen anywhere along the route between you and your remote site (in this case, the dslreports test site). The router isn't the only location. Cablemodems for instance will have their own buffers there.

The best way to get good result along most of the route (provided the ISP isn't overloaded) is to cap your bandwidth a few percents below what your ISP is giving you, ensuring that you will never fill up the buffers in your modem/router/PC.
I absolutely agree. It depends not only from your router settings, but also from concrete remote site. So if you tune your router for maximum performance with dslreports you can face problems with other sites.

I remember *DSL guys who used my FW set some linits for download/upload using QoS. There were some problems with QoS in my firmware due to lacking patch (missed in latest Netgear GPL codes). I succeed to restore it and he was satisfied by dslreports results.

Vladlenas: no sense to change something in your case. As far as I know you use cable connection (WAN).

Voxel.
 
I think that it was "PingPlotter" that I used to find latency bottlenecks along the internet path from me to the server on the other end when I was looking at bufferbloat. It was useful.
 

Latest threads

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top