no able to switch routing algo to V ?

Hello,

I just got openwrt 15.05.1 with batman-adv 2016.1 on 3 picostation HP, this time a selected before the build of openwrt batman V option, I can see it when it ask for available routing algo ,

root@OpenWrt:/etc/init.d# cat /sys/kernel/debug/batman_adv/routing_algos
Available routing algorithms:
  * BATMAN_IV
  * BATMAN_V

but when I do :

echo BATMAN_V > /sys/module/batman_adv/parameters/routing_algo

It does nothing, still on BATMAN_IV during session and on reboot ?

Quoted 11 lines · On Saturday 07 May 2016 16:08:35 contact wrote:

Hello,

I just got openwrt 15.05.1 with batman-adv 2016.1 on 3 picostation HP, this time a selected before the build of openwrt batman V option, I can see it when it ask for available routing algo ,

root@OpenWrt:/etc/init.d# cat /sys/kernel/debug/batman_adv/routing_algos
Available routing algorithms:
   * BATMAN_IV
   * BATMAN_V

Quoted 5 lines

but when I do :

echo BATMAN_V > /sys/module/batman_adv/parameters/routing_algo

It does nothing, still on BATMAN_IV during session and on reboot ?

This module parameter only controls what routing algorithm is chosen when a new batman-adv interface is created. So if you already have created it then changing the parameter doesn't do anything to the already existing batman-adv interface. Also rebooting a device will reset it either to the default or to the one you've explicitly specified during the loading of the modules.

But you seem to use OpenWrt. It allows to specify the routing_algo in /etc/config/network and should be applied right before the batman-adv interface is created. You just have to make sure that you don't specify conflicting things. An example of a simple configuration is:

config interface 'batnet'
option mtu '1532'
option proto 'batadv'
option mesh 'bat0'
option routing_algo 'BATMAN_V'

The wireless configuration would then normally reference this batnet network:

config wifi-iface 'wmesh'
option device 'radio0'
option ifname 'adhoc0'
option network 'batnet'
option mode 'adhoc'
option ssid 'mesh'
option 'mcast_rate' '18000'
option bssid '02:CA:FE:CA:CA:40'

This information was taken from Batman-adv-openwrt-config wiki page [1]

Kind regards,
Sven

[1] https://www.open-mesh.org/projects/batman-adv/wiki/Batman-adv-openwrt-config

Thank you, it is working and it is more clear for me.

I have switched to V but when I switch on other V nodes, my first node reset or freeze, probably a kernel panic. Need to plug a rs232 to be able to see what could be the problem.
I will switch back to IV.

Quoted 6 lines · On Saturday 07 May 2016 18:04:41 contact wrote:

Thank you, it is working and it is more clear for me.

I have switched to V but when I switch on other V nodes, my first node reset or freeze, probably a kernel panic. Need to plug a rs232 to be able to see what could be the problem.
I will switch back to IV.

Which version did you build?

https://github.com/openwrt-routing/packages/commit/8dec3dc38b514ed8f1212b51f9f6039af3fb0e04

or

https://github.com/openwrt-routing/packages/commit/a3cd8a1729779b8a39336a863ba7f97481e81f5c

or a different version/feed? Please make sure that you don't use a legacy feed together with batman-adv v2016.1 and B.A.T.M.A.N. V. It can only be used when build directly against the used mac80211 package which is only done in the two openwrt feeds linked here (and in a devel feed [1] which is building always against the current master).

Kind regards,
Sven

[1] https://git.open-mesh.org/openwrt-feed-devel.git

it was the default version present in yesterday openwrt stable 15.05.1

in batman makefile i have :

include $(TOPDIR)/rules.mk

PKG_NAME:=batman-adv

PKG_VERSION:=2016.1
PKG_RELEASE:=1
PKG_MD5SUM:=8c8e449009b4d29512d26ee308960bb5

Quoted 11 lines · On Saturday 07 May 2016 18:42:06 contact wrote:

it was the default version present in yesterday openwrt stable 15.05.1

in batman makefile i have :

include $(TOPDIR)/rules.mk

PKG_NAME:=batman-adv

PKG_VERSION:=2016.1
PKG_RELEASE:=1
PKG_MD5SUM:=8c8e449009b4d29512d26ee308960bb5

This doesn't say much (and you most likely meant openwrt-routing because openwrt itself doesn't provide batman-adv). But and Marek fixed a crash since yesterday. Not sure if it is related.

Kind regards,
Sven

yes it is in the routing feed I did not change anything from the openwrt files.

Ok, so you suggest to use the devel feed with thye last commit ?

Quoted 4 lines · On Saturday, May 07, 2016 18:51:40 contact wrote:

yes it is in the routing feed I did not change anything from the openwrt files.

Ok, so you suggest to use the devel feed with thye last commit ?

Please update your routing feed first. I pushed several stability updates today.

Thanks,
Marek

I just updated I need to recompile and re-flash my nodes it takes sometime, can you check please if all your patches are here ?

[..]

Fast-forward

  ...make-sure-ELP-OGM-orig-MAC-is-updated-on-.patch |  64 ++++++++++++
  ...-adv-Avoid-duplicate-neigh_node-additions.patch |  89 +++++++++++++++++
...Avoid-nullptr-derefence-in-batadv_v_neigh.patch |  38 +++++++
...an-adv-Fix-refcnt-leak-in-batadv_v_neigh_.patch |  81 +++++++++++++++
...Fix-double-neigh_node_put-in-batadv_v_ogm.patch |  40 ++++++++
...initialize-ELP-orig-address-on-secondary-.patch | 110

+++++++++++++++++++++

Quoted 5 lines · On Saturday, May 07, 2016 18:59:14 contact wrote:

I just updated I need to recompile and re-flash my nodes it takes sometime, can you check please if all your patches are here ?

[..]

Quoted 8 lines

Fast-forward

   ...make-sure-ELP-OGM-orig-MAC-is-updated-on-.patch |  64 ++++++++++++
   ...-adv-Avoid-duplicate-neigh_node-additions.patch |  89 +++++++++++++++++
...Avoid-nullptr-derefence-in-batadv_v_neigh.patch |  38 +++++++
...an-adv-Fix-refcnt-leak-in-batadv_v_neigh_.patch |  81 +++++++++++++++
...Fix-double-neigh_node_put-in-batadv_v_ogm.patch |  40 ++++++++
...initialize-ELP-orig-address-on-secondary-.patch | 110

+++++++++++++++++++++

Looks good. However, comparing git revisions (as Sven suggested) is more precise than posting such logs.

Cheers,
Marek

seems ok , I will give a try
(I came today at work only to do some mesh testing on moving robots :-) , never have the time during the week ...)

user@ubuntuvm:~/openwrt15011/openwrt$ git -C feeds/routing log
commit a3cd8a1729779b8a39336a863ba7f97481e81f5c
Author: Marek Lindner <[email protected]>
Date:   Sat May 7 20:42:44 2016 +0800

batman-adv: 2016.1 bugfixes & stability updates

Signed-off-by Marek Lindner <[email protected]>

commit 16ceff232dc58aad799bbb3d98915eb5c78ec2af

Author: Axel Neumann <[email protected]>
Date:   Fri May 6 10:35:47 2016 +0200

bmx7: bump version (report channel via topology plugin)

Quoted 22 lines · Marek Lindner

Quoted 12 lines · On Saturday, May 07, 2016 18:59:14 contact wrote:

I just updated I need to recompile and re-flash my nodes it takes sometime, can you check please if all your patches are here ?

[..]

Fast-forward

   ...make-sure-ELP-OGM-orig-MAC-is-updated-on-.patch |  64 
++++++++++++
   ...-adv-Avoid-duplicate-neigh_node-additions.patch |  89 
+++++++++++++++++
...Avoid-nullptr-derefence-in-batadv_v_neigh.patch |  38 +++++++
...an-adv-Fix-refcnt-leak-in-batadv_v_neigh_.patch |  81 

Quoted 4 lines

+++++++++++++++
...Fix-double-neigh_node_put-in-batadv_v_ogm.patch | 40 ++++++++ ...initialize-ELP-orig-address-on-secondary-.patch | 110
+++++++++++++++++++++

Looks good. However, comparing git revisions (as Sven suggested) is more precise than posting such logs.

Cheers,
Marek

I flashed 3 nodes (2 fixe and 1 mobile streaming 2mbits of video data) ,no crash for the moment.

I am making some test to see if all is ok.

Laurent

Quoted 44 lines · contact

seems ok , I will give a try
(I came today at work only to do some mesh testing on moving robots :-) , never have the time during the week ...)

user@ubuntuvm:~/openwrt15011/openwrt$ git -C feeds/routing log
commit a3cd8a1729779b8a39336a863ba7f97481e81f5c
Author: Marek Lindner <[email protected]>
Date:   Sat May 7 20:42:44 2016 +0800

batman-adv: 2016.1 bugfixes & stability updates

Signed-off-by Marek Lindner <[email protected]>

commit 16ceff232dc58aad799bbb3d98915eb5c78ec2af

Author: Axel Neumann <[email protected]>
Date:   Fri May 6 10:35:47 2016 +0200

bmx7: bump version (report channel via topology plugin)

Quoted 22 lines · Marek Lindner

Quoted 12 lines · On Saturday, May 07, 2016 18:59:14 contact wrote:

I just updated I need to recompile and re-flash my nodes it takes sometime, can you check please if all your patches are here ?

[..]

Fast-forward

   ...make-sure-ELP-OGM-orig-MAC-is-updated-on-.patch |  64 
++++++++++++
   ...-adv-Avoid-duplicate-neigh_node-additions.patch |  89 
+++++++++++++++++
...Avoid-nullptr-derefence-in-batadv_v_neigh.patch |  38 +++++++
...an-adv-Fix-refcnt-leak-in-batadv_v_neigh_.patch |  81 

Quoted 4 lines

+++++++++++++++
...Fix-double-neigh_node_put-in-batadv_v_ogm.patch | 40 ++++++++ ...initialize-ELP-orig-address-on-secondary-.patch | 110
+++++++++++++++++++++

Looks good. However, comparing git revisions (as Sven suggested) is more precise than posting such logs.

Cheers,
Marek

it seems table, I saw a really fast dynamic route creation. With batman IV the route took sometime 1 min to come and it was really noisy, i tried many settings,
here with V it is good, when I move toward a node it took less than 1 sec to create a new route and the video stream is not interrupted.

Sometime I loose the route it become a little bit noisy when I stop moving but here I suspect a wifi problem, need to change my channel.

I will make more test by adding other nodes on the path of the robot and see how routing is done.

yes it is in the routing feed I did not change anything from the openwrt files.

Ok, so you suggest to use the devel feed with thye last commit ?