commits

commits@lists.open-mesh.org

1 participants
12401 discussions

[batman-adv] master: batman-adv: bla: use netif_rx_ni when not in interrupt context (3747f81a)
by postmaster＠open-mesh.org 18 Aug '20

18 Aug '20

Repository : ssh://git@diktynna/batman-adv On branch : master >--------------------------------------------------------------- commit 3747f81a1380b65740fc52fc71c7a3af4c6e49de Author: Jussi Kivilinna <jussi.kivilinna(a)haltian.com> Date: Tue Aug 18 17:46:10 2020 +0300 batman-adv: bla: use netif_rx_ni when not in interrupt context batadv_bla_send_claim() gets called from worker thread context through batadv_bla_periodic_work(), thus netif_rx_ni needs to be used in that case. This fixes "NOHZ: local_softirq_pending 08" log messages seen when batman-adv is enabled. Fixes: a9ce0dc43e2c ("batman-adv: add basic bridge loop avoidance code") Signed-off-by: Jussi Kivilinna <jussi.kivilinna(a)haltian.com> Signed-off-by: Sven Eckelmann <sven(a)narfation.org> >--------------------------------------------------------------- 3747f81a1380b65740fc52fc71c7a3af4c6e49de net/batman-adv/bridge_loop_avoidance.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/batman-adv/bridge_loop_avoidance.c b/net/batman-adv/bridge_loop_avoidance.c index 41cc87f0..cfb9e16a 100644 --- a/net/batman-adv/bridge_loop_avoidance.c +++ b/net/batman-adv/bridge_loop_avoidance.c @@ -437,7 +437,10 @@ static void batadv_bla_send_claim(struct batadv_priv *bat_priv, u8 *mac, batadv_add_counter(bat_priv, BATADV_CNT_RX_BYTES, skb->len + ETH_HLEN); - netif_rx(skb); + if (in_interrupt()) + netif_rx(skb); + else + netif_rx_ni(skb); out: if (primary_if) batadv_hardif_put(primary_if);

1 0

[batman-adv] master: Merge branch 'maint' into master (9b44cc8b)
by postmaster＠open-mesh.org 18 Aug '20

18 Aug '20

Repository : ssh://git@diktynna/batman-adv On branch : master >--------------------------------------------------------------- commit 9b44cc8bc01ba21dcca431851d8ba4c0b421d65f Merge: eded19e9 3747f81a Author: Sven Eckelmann <sven(a)narfation.org> Date: Tue Aug 18 18:19:39 2020 +0200 Merge branch 'maint' into master >--------------------------------------------------------------- 9b44cc8bc01ba21dcca431851d8ba4c0b421d65f net/batman-adv/bridge_loop_avoidance.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)

1 0

[doc] master: Merge branch 'backup-redmine/2020-08-16' (db18301)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branch : master >--------------------------------------------------------------- commit db18301912b71a3e46588574bfe9a5393061950e Merge: 26ea6f6 20a0ea2 Author: Sven Eckelmann <sven(a)narfation.org> Date: Mon Aug 17 00:29:31 2020 +0200 Merge branch 'backup-redmine/2020-08-16' >--------------------------------------------------------------- db18301912b71a3e46588574bfe9a5393061950e batman-adv/Batman-adv-openwrt-config.rst | 70 +- .../Network-wide-multi-link-optimization.rst | 3 +- batman-adv/interface_detection.dia | 757 --------------------- batman-adv/interface_detection.svg | 57 -- batman-adv/primary_flag.dia | 595 ---------------- batman-adv/primary_flag.svg | 61 -- open-mesh/FFTrace.rst | 33 - open-mesh/Gsoc2010-ideas.rst | 211 ------ open-mesh/Gsoc2011-ideas.rst | 281 -------- open-mesh/Gsoc2011-roadmap.rst | 19 - open-mesh/Gsoc2012-ideas.rst | 14 - open-mesh/Kernel_hacking_Debian_image.rst | 1 + open-mesh/MeshLinux.rst | 29 - open-mesh/NetworkingTechnologies.rst | 42 -- open-mesh/News-draft.rst | 107 +-- open-mesh/Wbmv4-batmandev-agenda.rst | 34 - open-mesh/Wbmv5-batmandev-agenda.rst | 31 - open-mesh/Wbmv6-batmandev-agenda.rst | 68 -- open-mesh/WikiStart.rst | 7 - open-mesh/random.rst | 10 - 20 files changed, 38 insertions(+), 2392 deletions(-) diff --cc batman-adv/Batman-adv-openwrt-config.rst index 7168ee8,0000000..82dc90a mode 100644,000000..100644 --- a/batman-adv/Batman-adv-openwrt-config.rst +++ b/batman-adv/Batman-adv-openwrt-config.rst @@@ -1,320 -1,0 +1,254 @@@ +.. SPDX-License-Identifier: GPL-2.0 + +B.A.T.M.A.N. OpenWrt configuration +================================== + +This page showcases some basic batman-adv configurations on OpenWrt +including wireless & network setups. However, the goal is to provide not +more than a starting point. For a full reference of the OpenWrt uci +system please visit the official `OpenWrt uci +documentation <https://wiki.openwrt.org/doc/uci>`__. + +The configuration for the following common example is provided: One AP +with a wireless adhoc interface is supposed to run batman-adv +(essentially, the setup explained in our +:doc:`Quick-start-guide <Quick-start-guide>`). + +Batman-adv 2019.0-3 and newer +----------------------------- + +With batman-adv 2019.0-3, the OpenWrt package was modified to better integrate +in the netifd infrastructure. It now provides three different protos: + +* batadv_hardif + + * network interface used by batadv meshif to transport the batman-adv packets + * its master interface is set to the batadv meshif + +* batadv (meshif/softif) + + * virtual interface that emulates a normal 802.3 interface on top + * encapsulates traffic and forwards it via the batadv hardifs + +* batadv_vlan + + * potential VLAN ID on top of batadv meshif + * allows filtering of traffic from specific VIDs + +The wireless configuration in '/etc/config/wireless':: + + config wifi-device 'radio0' + option [..] + + config wifi-iface 'wmesh' + option device 'radio0' + option ifname 'mesh0' + option network 'bat0_hardif_mesh0' + option mode 'mesh' + option mesh_id 'mesh' + option mesh_fwding '0' + option mesh_ttl '1' + option 'mcast_rate' '24000' + + +It is assumed you configured the 'wifi-device' depending on your requirements +and your hardware. The interesting part is the 'wifi-iface' stanza with its +options: + +device + points back to your radio (wifi-device) interface + +ifname + allows you to specify an arbitrary name for your adhoc/meshpoint interface + +network + points to the corresponding stanza in '/etc/config/network' + +mode + defines the wifi mode - 802.11s mesh(point) in our case + +mcast_rate + helps to avoid low bandwidth routes (routes with a lower throughput rate than + the mcast rate will not be visible to batman-adv) + +mesh_id + is a basic wireless settings (like an SSID) you might want to set to + your liking + +More information can be found in the `OpenWrt wireless configuration <https://wiki.openwrt.org/doc/uci/wireless>`__ + +The first step is to create the "batadv" mesh interface (in our case "bat0") +in /etc/config/network with the optional list of options:: + + config interface 'bat0' + option proto 'batadv' + ## optional settings to override the defaults: + option routing_algo 'BATMAN_IV' + option aggregated_ogms 1 + option ap_isolation 0 + option bonding 0 + option fragmentation 1 + option gw_mode 'off' + #option gw_bandwidth '10000/2000' + #option gw_sel_class 20 + option log_level 0 + option orig_interval 1000 + option bridge_loop_avoidance 1 + option distributed_arp_table 1 + option multicast_mode 1 + option multicast_fanout 16 + option network_coding 0 + option hop_penalty 30 + option isolation_mark '0x00000000/0x00000000' + +The next step is to add actual network device has "batadv_hardif" to the "bat0" +batadv meshif. This is done by specifying a "batadv_hardif" interface section +per network device. Here we add eth0 and the mesh0 (from /etc/config/wireless) +to bat0. It is important to adjust the MTU of the batadv_hardif devices +to avoid fragmentation.:: + + # add *single* wifi-iface with network bat0_hardif_mesh0 as hardif to bat0 + config interface 'bat0_hardif_mesh0' + option proto 'batadv_hardif' + option master 'bat0' + option mtu '1536' + # option ifname is filled out by the wifi-iface + + # add eth0 as hardif to bat0 + config interface 'bat0_hardif_eth0' + option proto 'batadv_hardif' + option master 'bat0' + option mtu '1536' + option ifname 'eth0' + option 'elp_interval' 500 + # change throughput_override to 0 to use automatic detection; also allows kbit suffix + option 'throughput_override' '1mbit' + +The "bat0" batadv meshif can then be used like any other network device. It +can be added to bridges by adding "bat0" to the list of "ifnames" of the bridge. +Or an IP can be configured using:: + + # configure IP on bat0 + config interface 'bat0_lan' + option ifname 'bat0' + option proto 'static' + option ipaddr '192.168.1.1' + option netmask '255.255.255.0' + option ip6assign '60' + +VLAN specific options have to be configured in a separated stanza having +protocol 'batadv_vlan'. + +In this particular section the user has to specify the name of the VLAN +interface (that will be automatically created by netifd) and then list all the +wanted options. At the moment the only available option for this section is +'ap_isolation':: + + config interface 'my_bat_vlan1' + option proto 'batadv_vlan' + option ifname 'bat0.1' + option ap_isolation '1' + +Any VLAN specific option can still be listed in the batadv meshif configuration +for 'bat0' and in that case they will affect the behaviour only for the plain +mesh interface (i.e. bat0 - the untagged LAN). + + - Batman-adv 2013.0.0 and till 2019.0-2 - -------------------------------------- ++Batman-adv 2019.0-2 and older ++------------------------------ + +With batman-adv 2013.0.0 the OpenWrt package was converted to integrate +with OpenWrt's netifd system. This has some impact on the configuration +but not much. + +The wireless configuration in '/etc/config/wireless': + +:: + + config wifi-device 'radio0' + option [..] + + config wifi-iface 'wmesh' + option device 'radio0' + option ifname 'adhoc0' + option network 'bat0_hardif_wlan' + option mode 'adhoc' + option ssid 'mesh' + option 'mcast_rate' '18000' + option bssid '02:CA:FE:CA:CA:40' + +It is assumed you configured the 'wifi-device' depending on your +requirements and your hardware. The interesting part is the 'wifi-iface' +stanza with its options: + +* 'device' points back to your radio (wifi-device) interface +* 'ifname' allows you to specify an arbitrary name for your adhoc + interface (which we are going to re-use later) +* 'network' points to the corresponding stanza in + '/etc/config/network' +* 'mode' defines the wifi mode (adhoc in our case) +* 'mcast\_rate' helps to avoid low bandwidth routes (routes with a + lower throughput rate than the mcast rate will not be visible to + batman-adv) +* 'ssid' and 'bssid' are basic wireless settings you might want to + set to your liking + +More information can be found in the `OpenWrt wireless +configuration <https://wiki.openwrt.org/doc/uci/wireless>`__ + +The network configuration in '/etc/config/network' allows to specify +batman-adv as protocol and the batman-adv interface name. + +:: + + config interface 'bat0_hardif_wlan' + option mtu '1532' + option proto 'batadv' + option mesh 'bat0' + + config interface 'bat0_hardif_eth0' + option mtu '1532' + option proto 'batadv' + option mesh 'bat0' + option ifname 'eth0' + +The batman-adv configuration in '/etc/config/batman-adv' only contains +the batman-adv specific options: + +:: + + config mesh 'bat0' + option 'aggregated_ogms' + option 'ap_isolation' + option [..] + +**Since batman-adv 2014.2.0** VLAN specific options have to be +configured in a separated stanza having protocol 'batadv\_vlan'. + +In this particular section the user has to specify the name of the VLAN +interface (that will be automatically created by netifd) and then list +all the wanted options. At the moment the only available option for this +section is 'ap\_isolation': + +:: + + config interface 'my_bat_vlan1' + option proto 'batadv_vlan' + option ifname 'bat0.1' + option ap_isolation '1' + +Any VLAN specific option can still be listed in the batman-adv +configuration for 'bat0' and in that case they will affect the behaviour +only for the plain mesh interface (i.e. bat0 - the untagged LAN). + +**Since batman-adv 2016.1** The routing algorithm can be configured via +the network configuration in '/etc/config/network' to override the +kernel module's default: + +:: + + config interface 'bat0_hardif_wlan' + option mtu '1532' + option proto 'batadv' + option mesh 'bat0' + option routing_algo 'BATMAN_V' - - Batman-adv 2012.4.0 and older - ----------------------------- - - The wifi interface needs to be configured in '/etc/config/wireless': - - :: - - config wifi-device 'radio0' - option [..] - - config wifi-iface 'wmesh' - option device 'radio0' - option ifname 'adhoc0' - option network 'bat0_hardif_wlan' - option mode 'adhoc' - option ssid 'mesh' - option 'mcast_rate' '18000' - option bssid '02:CA:FE:CA:CA:40' - - It is assumed you configured the 'wifi-device' depending on your - requirements and your hardware. The interesting part is the 'wifi-iface' - stanza with its options: - - * 'device' points back to your radio (wifi-device) interface - * 'ifname' allows you to specify an arbitrary name for your adhoc - interface (which we are going to re-use later) - * 'network' points to the corresponding stanza in - '/etc/config/network' - * 'mode' defines the wifi mode (adhoc in our case) - * 'mcast\_rate' helps to avoid low bandwidth routes (routes with a - lower throughput rate than the mcast rate will not be visible to - batman-adv) - * 'ssid' and 'bssid' are basic wireless settings you might want to - set to your liking - - More information can be found in the `OpenWrt wireless - configuration <https://wiki.openwrt.org/doc/uci/wireless>`__ - - Configure the MTU in '/etc/config/network' - - :: - - config interface 'bat0_hardif_wlan' - option ifname 'adhoc0' - option mtu '1528' - option proto 'none' - - The stanza's name 'mesh' as well as the ifname option have to match your - wireless configuration. - - Batman-adv is configured through its own configuration in - '/etc/config/batman-adv': - - :: - - config mesh 'bat0' - option interfaces 'adhoc0' - option 'aggregated_ogms' - option 'ap_isolation' - option [..] - - The 'interfaces' option is the key element here, as it tells batman-adv - which interface(s) to run on. All the other options enable / disable / - tweak all the batman-adv features you can also access at runtime through - batctl. diff --cc batman-adv/Network-wide-multi-link-optimization.rst index 62dde4b,0000000..0f72272 mode 100644,000000..100644 --- a/batman-adv/Network-wide-multi-link-optimization.rst +++ b/batman-adv/Network-wide-multi-link-optimization.rst @@@ -1,177 -1,0 +1,176 @@@ +.. SPDX-License-Identifier: GPL-2.0 + +Network Wide Multi Link Optimization (technical documentation) +============================================================== + - The original batman-adv multi link optimization - (:doc:`Multi-link-optimizations-technical <Multi-link-optimize>`) works on a local level - which ++The original batman-adv multi link optimization worked on a local level - which +is fairly easy to implement and works well as long as as all nodes are +configured the same way and have similar link qualities. But local +decision may be suboptimal network wide - for example, consider some +dual radio mesh nodes where the first node starts on 5 GHz, but after +some hops (alternating between 2.4 and 5 GHz) it stops at a 2.4 GHz-only +device. If it had chosen the other (2.4 Ghz) link in the first hop, +choosing the same frequency twice could have been avoided, and the total +path quality would be better. + +|image0| + +This limitation can be solved by relaying the information about +multi-interfaces over the whole mesh. This requires some changes to +routing and forwarding, which are to be described here. + +Routing Table Per Interface +--------------------------- + +The main idea is to use n + 1 routing tables instead of just one, with n +being the number of configured interfaces: + +* a routing table per incoming interface +* an additional 'default' routing table for packets generated locally + (e.g. from the soft-interface) + +The routing table is chosen based on the interface the packet entered +the system by. + +OGM forwarding and penalties +---------------------------- + +To propagate different paths, the OGM forwarding is changed: + +* When there are multiple interfaces and an OGM is forwarded on the + same interface, an additional penalty (e.g. "half duplex penalty) when + re-forwarding the OGM on the same interface + -> the metric with the **applied** penalty is stored locally +* apply the strict "forward only from best neighbor" per interface + +*Path Metric computation* +Each node, creates and sends its own OGMs to let all the other peers +in the network build their routing table + +- on each interface, one OGM is sent as broadcast containing the + initial metric value + +- A generic node receiving an OGM from a neighbor on interface I. When + forwarding, it will consider the output interface O: + + - perform the usual B.A.T.M.A.N. metric computations, including + accepting OGMs only from the best neighbor + - compare incoming interface I and outgoing interface O, and apply + special rules based on that (e.g. a penalty if I and O are the + same, and the interface is a wifi interface) + - mark the computed metric in a special originator table for the + interface O + - forward the OGM with this computed metric only on interface O + +- all metric computations above are also performed for a virtual + 'default' interface O, which is used for local traffic. + +*Traffic forwarding* + +- traffic forwarding works in the reverse direction: + + - When forwarding traffic, use the originator table of the interface + where the packet was received + - for locally generated traffic, use the default originator table + +For example, consider the following illustration: + +|image1| + +We consider all link qualities equal. What should happen according to +the rules (after some OGMs): + +* OGMs from A are sent via A1 and A2 +* B applies the new penalty when forwarding packets from A1 again on + B1, and also when forwarding from A2 on B2 +* therefore, As OGM received on B1 are only forwarded on B2 (because + no penalty is applied here). They are not re-forwarded on B1 because + for this interface, the packets received on B2 have a better quality. +* finally (assuming equal link qualities everywhere), the TQ values + in the OGMs from A forwarded by B are equal too - but the internal + routing table include the interface switching +* putting it all together and assuming the other nodes behave the + same way: If a packet is sent by D on D1, it will follow the dashed + path. If it is sent on D2, it will follow the solid path. The + interface alternation was implemented by local routing table decisions + but forwarded information (unlike the original local-only interface + alternation). + +Another example: path diversity +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +|image2| + +Scenario: + +- Node B is equipped with 2 interfaces (imagine two wifi cards with + sectored antennas) +- Node A is connected to B through B1 and uses B as next hop towards E +- Node F is connected to B through B2 and uses B as next hop towards E +- All the links are perfect +- Packets sent by A intended to E are received by B on B1 and forwarded + using B2 +- Packets sent by F intended to E are received by B on B2 and forwarded + using B1 +- B's nexthop to E using B1 is different from B's nexthop to E using B2 + +The presented example is another consequence of the mechanism +explained in this page. +Since the multi-interface optimisation is not per-link anymore but is +now defined network wide, **packets can possibly be routed through two +completely different paths in order to reach the same destination**. + +This possibility is given by the fact that on the path from A to E B +will choose G as best next-hop, while on the path from F to E B will +choose C. + +The result is that, with this new feature, multi-interface nodes can act +as routing splitting point which leads to a first tempative of real +multi-path routing. + +Debug information +----------------- + +There is an implementation based on the description above currently +pending for review. It offers some debugging facilities: + +The default originator table can be found at its original place, using +''batctl o''. The following examples represents the topology shown in the top +of the page, debug information obtained on node B, using OpenWRT and +virtual machines: + +:: + + root@OpenWrt:/# batctl o + [B.A.T.M.A.N. adv master-b82b9b2, MainIF/MAC: eth0/fe:f0:00:00:02:01 (bat0 BATMAN_IV)] + Originator last-seen (#/255) Nexthop [outgoingIF]: Potential nexthops ... + fe:f1:00:00:03:01 0.350s (254) fe:f1:00:00:03:01 [ eth1]: fe:f1:00:00:03:01 (254) + fe:f1:00:00:01:01 0.800s (255) fe:f1:00:00:01:01 [ eth1]: fe:f1:00:00:01:01 (255) + fe:f0:00:00:05:01 0.770s (225) fe:f1:00:00:03:01 [ eth1]: fe:f0:00:00:03:01 (211) fe:f1:00:00:03:01 (225) + fe:f0:00:00:03:01 0.670s (255) fe:f0:00:00:03:01 [ eth0]: fe:f1:00:00:03:01 (255) fe:f0:00:00:03:01 (255) + fe:f0:00:00:04:01 0.520s (234) fe:f1:00:00:03:01 [ eth1]: fe:f1:00:00:03:01 (234) fe:f0:00:00:03:01 (222) + fe:f0:00:00:01:01 0.920s (255) fe:f1:00:00:01:01 [ eth1]: fe:f1:00:00:01:01 (255) fe:f0:00:00:01:01 (254) + +Comparing to the original topology, it shows that for destination E +(fe:f0:00:00:05:01) when packets come in on 2.4 GHz (eth0, first table), +they are preferably forwarded to C (fe:f1:00:00:03:01 ) on 5 GHz (eth1). + +When packets are received on 5 GHz (eth1), there is not much difference +because at node B or at node C, the same interface must be used. +Therefore in the second table for node E (fe:f0:00:00:05:01) the two +choices have pretty much the same TQ values (210 and 211). + +Ideas for the future +-------------------- + +- Theoretically, a node generating traffic instead of using its own + routing table could exploit the path diversity and route its traffic + using all the routing tables of all the interfaces (the selection + policy have to be defined..round robin would not work properly + because when using two different paths TCP reordering may decrease + the performance) + +.. |image0| image:: alternating-limited-view.svg +.. |image1| image:: alternation_chain.svg +.. |image2| image:: net-wide-multiif.svg + diff --cc open-mesh/Kernel_hacking_Debian_image.rst index b03350d,0000000..2ac30e3 mode 100644,000000..100644 --- a/open-mesh/Kernel_hacking_Debian_image.rst +++ b/open-mesh/Kernel_hacking_Debian_image.rst @@@ -1,440 -1,0 +1,441 @@@ +.. SPDX-License-Identifier: GPL-2.0 + +Kernel hacking Debian image +=========================== + +The :doc:`OpenWrt image <OpenWrt_in_QEMU>` is an easy way to start multiple +virtual instances. But these instances usually don��t provide the +required infrastructure to test kernel modules extensively. And it also +depends on special toolchains to prepare the used tools/modules which +should tested. + +It is often easier to use the same operating system in the virtual +environment and on the host. Only the kernel is modified here to provide +the necessary helpers for in-kernel development. + +An interested reader might even extend this further to only provide a +modified kernel and use the currently running rootfs also in the virtual +environment. Such an approach is used in `hostap��s test +vm <https://w1.fi/cgit/hostap/tree/tests/hwsim/vm>`__ but it is out of +scope for this document. + +Create an Image +--------------- + +The debian root filesystem is used here to a minimal system to boot and +run the test programs. It is a simple ext4 filesystem with only +userspace components from Debian. The configuration is changed to: + +* automatically mount the shared folder +* automatically set up a static IPv4 address and hostname on bootup +* start a test-init.sh script from the shared folder on bootup +* disable root password +* prefer batctl binary from shared folder��s batctl subdirectory instead + of virtual environment binary + +The installation is also cleaned up at the end to reduce the required +storage space + +.. code-block:: sh + + qemu-img create debian.img 8G + sudo mkfs.ext4 -O '^has_journal' -F debian.img + sudo mkdir debian + sudo mount -o loop debian.img debian + sudo debootstrap buster debian + sudo chroot debian apt update + sudo chroot debian apt install --no-install-recommends build-essential vim openssh-server less \ + pkg-config libnl-3-dev libnl-genl-3-dev libcap-dev tcpdump rng-tools5 \ + trace-cmd flex bison libelf-dev libdw-dev binutils-dev libunwind-dev libssl-dev libslang2-dev liblzma-dev libperl-dev + + sudo mkdir debian/root/.ssh/ + ssh-add -L | sudo tee debian/root/.ssh/authorized_keys + + sudo mkdir debian/host + sudo sh -c 'cat > debian/etc/fstab << EOF + host /host 9p trans=virtio,version=9p2000.L 0 0 + EOF' + + sudo sh -c 'cat > debian/etc/rc.local << "EOF" + #!/bin/sh -e + + MAC_PART="$(ip link show enp0s1 | awk "/ether/ {print \$2}"| sed -e "s/.*://" -e "s/[\\n\\ ].*//"|awk "{print (\"0x\"\$1)*1 }")" + IP_PART="$(echo $MAC_PART|awk "{ print \$1+50 }")" + NODE_NR="$(echo $MAC_PART|awk "{ printf(\"%02d\", \$1) }")" + ip addr add 192.168.251.${IP_PART}/24 dev enp0s1 + ip link set up dev enp0s1 + hostname "node"$NODE_NR + ip link set up dev lo + [ ! -x /host/test-init.sh ] || /host/test-init.sh + exit 0 + EOF' + sudo chmod a+x debian/etc/rc.local + + sudo sed -i 's/^root:[^:]*:/root::/' debian/etc/shadow + + ## optionally: allow ssh logins without passwords + # sudo sed -i 's/^#PermitRootLogin.*/PermitRootLogin yes/' debian/etc/ssh/sshd_config + # sudo sed -i 's/^#PermitEmptyPasswords.*/PermitEmptyPasswords yes/' debian/etc/ssh/sshd_config + # sudo sed -i 's/^UsePAM.*/UsePAM no/' debian/etc/ssh/sshd_config + + ## optionally: enable autologin for user root + #sudo mkdir debian/etc/systemd/system/serial-getty(a)hvc0.service.d/ + #sudo sh -c 'cat > debian/etc/systemd/system/serial-getty(a)hvc0.service.d/autologin.conf << EOF + #[Service] + #ExecStart= + #ExecStart=-/sbin/agetty --autologin root -s %I 115200,38400,9600 vt102 + #EOF' + + sudo sh -c 'echo '\''PATH="/host/batctl/:$PATH"'\'' >> debian/etc/profile' + sudo rm debian/var/cache/apt/archives/*.deb + sudo rm debian/var/lib/apt/lists/* + sudo e4defrag -v debian/ + sudo umount debian + sudo fsck.ext4 -fD debian.img + sudo zerofree -v debian.img + sudo fallocate --dig-holes debian.img + + + ## optionally: convert image to qcow2 + #sudo qemu-img convert -c -f raw -O qcow2 debian.img debian.qcow2 + #sudo mv debian.qcow2 debian.img + +Kernel compile +-------------- + +Any recent kernel can be used for the setup. We will use linux-next here +to get the most recent development kernels. It is also assumed that the +sources are copied to the same directory as the debian.img and a x86_64 +image will be used. + +The kernel will be build to enhance the virtualization and debugging +experience. It is configured with: + +* basic kernel features +* support for necessary drivers +* kernel hacking helpers +* kernel address + undefined sanitizers +* support for hwsim + +.. code-block:: sh + + # make sure that libelf-dev is installed or module build will fail with something like "No rule to make target 'net/batman-adv/bat_algo.o'" + + git clone git://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git + cd linux-next + + make allnoconfig + cat >> .config << EOF + + # small configuration + CONFIG_SMP=y + CONFIG_EMBEDDED=n + # CONFIG_EXPERT is not set + CONFIG_MODULES=y + CONFIG_MODULE_UNLOAD=y + CONFIG_MODVERSIONS=y + CONFIG_MODULE_SRCVERSION_ALL=y + CONFIG_64BIT=y + CONFIG_X86_VSYSCALL_EMULATION=n + CONFIG_IA32_EMULATION=n + CONFIG_VOP_BUS=y + CONFIG_VOP=y + CONFIG_HW_RANDOM_VIRTIO=y + CONFIG_NET_9P_VIRTIO=y + CONFIG_VIRTIO_MENU=y + CONFIG_SCSI_VIRTIO=y + CONFIG_VIRTIO_BALLOON=y + CONFIG_VIRTIO_BLK=y + CONFIG_VIRTIO_CONSOLE=y + CONFIG_VIRTIO_INPUT=y + CONFIG_VIRTIO_NET=y + CONFIG_VIRTIO_PCI=y + CONFIG_VIRTIO_PCI_LEGACY=y + CONFIG_VIRTIO_BALLOON=y + CONFIG_VIRTIO_BLK_SCSI=y + CONFIG_VIRTIO_INPUT=y + CONFIG_VIRTIO_MMIO=y + CONFIG_VIRTIO_MMIO_CMDLINE_DEVICES=y + CONFIG_RPMSG_VIRTIO=y + CONFIG_VSOCKETS=y + CONFIG_VIRTIO_VSOCKETS=y + CONFIG_DRM=y + CONFIG_DRM_VIRTIO_GPU=y + CONFIG_CAIF=y + CONFIG_CAIF_VIRTIO=y + CONFIG_CRYPTO_DEV_VIRTIO=y + CONFIG_FUSE_FS=y + CONFIG_VIRTIO_FS=y + CONFIG_IOMMU_SUPPORT=y + CONFIG_VIRTIO_IOMMU=y + CONFIG_LIBNVDIMM=y + CONFIG_VIRTIO_PMEM=y + CONFIG_CRC16=y + CONFIG_LIBCRC32C=y + CONFIG_CRYPTO_SHA512=y + CONFIG_NET=y + CONFIG_INET=y + CONFIG_DEBUG_FS=y + CONFIG_IPV6=y + CONFIG_BRIDGE=y + CONFIG_VLAN_8021Q=y + CONFIG_WIRELESS=n + CONFIG_NET_9P=y + CONFIG_NETWORK_FILESYSTEMS=y + CONFIG_9P_FS=y + CONFIG_9P_FS_POSIX_ACL=y + CONFIG_9P_FS_SECURITY=y + CONFIG_BLOCK=y + CONFIG_BLK_DEV=y + CONFIG_EXT4_FS=y + CONFIG_EXT4_USE_FOR_EXT23=y + CONFIG_TTY=y + CONFIG_HW_RANDOM=y + CONFIG_VHOST_RING=y + CONFIG_GENERIC_ALLOCATOR=y + CONFIG_SCSI_LOWLEVEL=y + CONFIG_SCSI=y + CONFIG_NETDEVICES=y + CONFIG_NET_CORE=y + CONFIG_DEVTMPFS=y + CONFIG_HYPERVISOR_GUEST=y + CONFIG_PVH=y + CONFIG_PARAVIRT=y + CONFIG_PARAVIRT_TIME_ACCOUNTING=y + CONFIG_PARAVIRT_SPINLOCKS=y + CONFIG_KVM_GUEST=y + CONFIG_BINFMT_ELF=y + CONFIG_BINFMT_SCRIPT=y + CONFIG_BINFMT_MISC=y + CONFIG_PCI=y + CONFIG_SYSVIPC=y + CONFIG_POSIX_MQUEUE=y + CONFIG_CROSS_MEMORY_ATTACH=y + CONFIG_UNIX=y + CONFIG_TMPFS=y + CONFIG_CGROUPS=y + CONFIG_BLK_CGROUP=y + CONFIG_CGROUP_CPUACCT=y + CONFIG_CGROUP_DEVICE=y + CONFIG_CGROUP_FREEZER=y + CONFIG_CGROUP_HUGETLB=y + CONFIG_CGROUP_NET_CLASSID=y + CONFIG_CGROUP_NET_PRIO=y + CONFIG_CGROUP_PERF=y + CONFIG_CGROUP_SCHED=y + CONFIG_DEVPTS_MULTIPLE_INSTANCES=y + CONFIG_INOTIFY_USER=y + CONFIG_FHANDLE=y + CONFIG_CFG80211=y + CONFIG_DUMMY=y + CONFIG_PACKET=y + CONFIG_VETH=y + CONFIG_IP_MULTICAST=y + CONFIG_NET_IPGRE_DEMUX=y + CONFIG_NET_IP_TUNNEL=y + CONFIG_NET_IPGRE=y + CONFIG_NET_IPGRE_BROADCAST=y + # CONFIG_LEGACY_PTYS is not set + CONFIG_NO_HZ_IDLE=y + CONFIG_CPU_IDLE_GOV_HALTPOLL=y + CONFIG_PVPANIC=y + + # makes boot a lot slower but required for shutdown + CONFIG_ACPI=y + + + #debug stuff + CONFIG_CC_STACKPROTECTOR_STRONG=y + CONFIG_LOCKUP_DETECTOR=y + CONFIG_DETECT_HUNG_TASK=y + CONFIG_SCHED_STACK_END_CHECK=y + CONFIG_DEBUG_RT_MUTEXES=y + CONFIG_DEBUG_SPINLOCK=y + CONFIG_DEBUG_MUTEXES=y + CONFIG_PROVE_LOCKING=y + CONFIG_LOCK_STAT=y + CONFIG_DEBUG_LOCKDEP=y + CONFIG_DEBUG_ATOMIC_SLEEP=y + CONFIG_DEBUG_LIST=y + CONFIG_DEBUG_PI_LIST=y + CONFIG_DEBUG_SG=y + CONFIG_DEBUG_NOTIFIERS=y + CONFIG_PROVE_RCU_REPEATEDLY=y + CONFIG_SPARSE_RCU_POINTER=y + CONFIG_DEBUG_STRICT_USER_COPY_CHECKS=y + CONFIG_X86_VERBOSE_BOOTUP=y + CONFIG_DEBUG_RODATA=y + CONFIG_DEBUG_RODATA_TEST=n + CONFIG_DEBUG_SET_MODULE_RONX=y + CONFIG_PAGE_EXTENSION=y + CONFIG_DEBUG_PAGEALLOC=y + CONFIG_DEBUG_OBJECTS=y + CONFIG_DEBUG_OBJECTS_FREE=y + CONFIG_DEBUG_OBJECTS_TIMERS=y + CONFIG_DEBUG_OBJECTS_WORK=y + CONFIG_DEBUG_OBJECTS_RCU_HEAD=y + CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER=y + CONFIG_DEBUG_KMEMLEAK=y + CONFIG_DEBUG_KMEMLEAK_EARLY_LOG_SIZE=8000 + CONFIG_DEBUG_STACK_USAGE=y + CONFIG_DEBUG_STACKOVERFLOW=y + CONFIG_DEBUG_INFO=y + CONFIG_DEBUG_INFO_DWARF4=y + CONFIG_GDB_SCRIPTS=y + CONFIG_READABLE_ASM=y + CONFIG_STACK_VALIDATION=y + CONFIG_WQ_WATCHDOG=y + CONFIG_DEBUG_KOBJECT_RELEASE=y + CONFIG_DEBUG_WQ_FORCE_RR_CPU=y + CONFIG_OPTIMIZE_INLINING=y + CONFIG_ENABLE_MUST_CHECK=y + CONFIG_ENABLE_WARN_DEPRECATED=y + CONFIG_DEBUG_SECTION_MISMATCH=y + CONFIG_UNWINDER_ORC=y + CONFIG_FTRACE=y + CONFIG_FUNCTION_TRACER=y + CONFIG_FUNCTION_GRAPH_TRACER=y + CONFIG_FTRACE_SYSCALLS=y + CONFIG_TRACER_SNAPSHOT=y + CONFIG_TRACER_SNAPSHOT_PER_CPU_SWAP=y + CONFIG_STACK_TRACER=y + CONFIG_UPROBE_EVENTS=y + CONFIG_DYNAMIC_FTRACE=y + CONFIG_FUNCTION_PROFILER=y + CONFIG_HIST_TRIGGERS=y + CONFIG_SYMBOLIC_ERRNAME=y + CONFIG_DYNAMIC_DEBUG=y + CONFIG_PRINTK_TIME=y + CONFIG_PRINTK_CALLER=y + CONFIG_DEBUG_MISC=y + CONFIG_PROVE_RCU_LIST=y ++ CONFIG_DEBUG_FORCE_FUNCTION_ALIGN_32B=y + + # for GCC 5+ + CONFIG_KASAN=y + CONFIG_KASAN_INLINE=y + CONFIG_UBSAN_SANITIZE_ALL=y + CONFIG_UBSAN=y + CONFIG_UBSAN_NULL=y + EOF + make olddefconfig + + cat >> .config << EOF + # allow to use unsigned regdb with hwsim + CONFIG_EXPERT=y + CONFIG_CFG80211_CERTIFICATION_ONUS=y + # CONFIG_CFG80211_REQUIRE_SIGNED_REGDB is not set + EOF + make olddefconfig + + make all -j$(nproc || echo 1) + +Build the BIOS +-------------- + +The (sea)bios used by qemu is nice to boot all kind of legacy images but +reduces the performance for booting a paravirtualized Linux system. +Something like qboot works better for this purpose: + +.. code-block:: sh + + git clone https://github.com/bonzini/qboot.git + cd qboot + meson build && ninja -C build + cd .. + +.. _open-mesh-kernel-hacking-debian-image-building-the-batman-adv-module: + +Building the batman-adv module +------------------------------ + +The kernel module can be build outside the virtual environment and +shared over the 9p mount. The path to the kernel sources have to be +provided to the make process + +.. code-block:: sh + + make KERNELPATH="$(pwd)/../linux-next" + +The kernel module can also be compiled in a way which creates better +stack traces and increases the usability with (k)gdb: + +.. code-block:: sh + + make EXTRA_CFLAGS="-fno-inline -Og -fno-optimize-sibling-calls" KERNELPATH="$(pwd)/../linux-next" V=1 + +Start of the environment +------------------------ + +virtual network initialization +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +The +:ref:`virtual-network.sh from the OpenWrt environment <open-mesh-openwrt-in-qemu-virtual-network-initialization>` +can be reused again. + +VM instances bringup +~~~~~~~~~~~~~~~~~~~~ + +The +:ref:`run.sh from the OpenWrt environment <open-mesh-openwrt-in-qemu-vm-instances-bringup>` +can mostly be reused. There are only minimal +adjustments required. + +The BASE_IMG is of course no longer the same because a new image +��debian.img�� was created for our new environment. The image also doesn��t +contain a bootloader or kernel anymore. The kernel must now be supplied +manually to qemu. + +.. code-block:: sh + + BASE_IMG=debian.img + BOOTARGS+=("-bios" "qboot/build/bios.bin") + BOOTARGS+=("-kernel" "linux-next/arch/x86/boot/bzImage") + BOOTARGS+=("-append" "root=/dev/vda rw console=hvc0 nokaslr tsc=reliable no_timer_check noreplace-smp rootfstype=ext4 rcupdate.rcu_expedited=1 reboot=t pci=lastbus=0 i8042.direct=1 i8042.dumbkbd=1 i8042.nopnp=1 i8042.noaux=1") + BOOTARGS+=("-device" "virtconsole,chardev=charconsole0,id=console0") + +It is also recommended to use linux-next/vmlinux instead of bzImage with +qemu 4.0.0 (or later) + +Automatic test initialization +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +The +:ref:`test-init.sh from the OpenWrt environment <open-mesh-openwrt-in-qemu-automatic-test-initialization>` +is always test specific. But its main +functionality is still the same as before. A simple example would be: + +.. code-block:: sh + + cat > test-init.sh << "EOF" + #! /bin/sh + + set -e + + ## get internet access + dhclient enp0s2 + + ## Simple batman-adv setup + + # ip link add dummy0 type dummy + ip link set up dummy0 + + rmmod batman-adv || true + insmod /host/batman-adv/net/batman-adv/batman-adv.ko + /host/batctl/batctl routing_algo BATMAN_IV + /host/batctl/batctl if add dummy0 + /host/batctl/batctl it 5000 + /host/batctl/batctl if add enp0s1 + ip link set up dev enp0s1 + ip link set up dev bat0 + EOF + + chmod +x test-init.sh + +Start +----- + +The startup method +:ref:`from the OpenWrt environment <open-mesh-openwrt-in-qemu-start>` +should be used here. diff --cc open-mesh/News-draft.rst index 47e597c,0000000..008e612 mode 100644,000000..100644 --- a/open-mesh/News-draft.rst +++ b/open-mesh/News-draft.rst @@@ -1,155 -1,0 +1,116 @@@ +.. SPDX-License-Identifier: GPL-2.0 + - DRAFT: Batman-adv 2020.2 released ++DRAFT: Batman-adv 2020.3 released +================================= + - Jul 1th, 2020. Today the B.A.T.M.A.N. team publishes the July 2020 - update to batman-adv, batctl and alfred! This release drops the support - for the end-of-life Linux kernel versions (<4.3). The B.A.T.M.A.N. V - throughput detection for ethernet devices without auto-negotiation was - enabled again to support more configurations out of the box. Most users - will still require a manual setting via the throughput_override for such - devices. The automatic removal of (empty) batadv mesh interfaces during - a batctl meshif bat0 interface del .... was dropped. An explicit batctl - meshif bat0 interface destroy is now required to remove the batadv - interface. Also several bugfixes and code cleanups are included in this - version. ++Aug 25th, 2020. Today the B.A.T.M.A.N. team publishes the August 2020 ++update to batman-adv, batctl and alfred! An additional hop penalty can ++now be configured on a per interface basis. Also several bugfixes and ++code cleanups are included in this version. + +As the kernel module always depends on the Linux kernel it is compiled +against, it does not make sense to provide binaries on our website. As +usual, you will find the signed tarballs in our download section: + - https://downloads.open-mesh.org/batman/releases/batman-adv-2020.2/ ++https://downloads.open-mesh.org/batman/releases/batman-adv-2020.3/ + +Thanks +------ + +Thanks to all people sending in patches: + - * Al Viro <viro(a)zeniv.linux.org.uk> - * Antonio Quartulli <a(a)unstable.cc> - * Cong Wang <xiyou.wangcong(a)gmail.com> - * Marek Lindner <mareklindner(a)neomailbox.ch> +* Linus L��ssing <linus.luessing(a)c0d3.blue> +* Simon Wunderlich <sw(a)simonwunderlich.de> +* Sven Eckelmann <sven(a)narfation.org> - * YueHaibing <yuehaibing(a)huawei.com> + +and to all those that supported us with good advice or rigorous testing: + - * Andreas Ziegler <az(a)freifunk-altdorf.de> - * Matthias Schiffer <mschiffer(a)universe-factory.net> - * Taehee Yoo <ap420073(a)gmail.com> ++* Antonio Quartulli <a(a)unstable.cc> + +batman-adv +---------- + +:: + + $ git describe origin/master - v2020.1-13-ga042622f - $ range=v2020.1..v2020.1-13-ga042622f ++ v2020.2-7-geded19e9 ++ $ range=v2020.2..v2020.2-7-geded19e9 + $ git shortlog --email --no-merges "${range}" + $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq + + - new kernel support - ================== - - batman-adv: Drop support for Linux < 4.4 - + coding style cleanup/refactoring + ================================ + - batman-adv: get rid of pointless access_ok() - batman-adv: use rcu_replace_pointer() where appropriate - batman-adv: partially revert dynamic lockdep key changes - batman-adv: change addr_list_lock back to static key - batman-adv: Fix spelling error in term buffer - batman-adv: trace: Drop unneeded types.h include - batman-adv: Utilize prandom_u32_max for random [0, max) values - batman-adv: Sync config options for external module builds - batman-adv: Use alphabetical order for external build config options ++ batman-adv: Switch mailing list subscription page ++ batman-adv: Fix typos and grammar in documentation + - BATMAN V - ========= ++ various ++ ======= + - batman-adv: Revert "disable ethtool link speed detection when auto negotiation off" ++ batman-adv: Introduce a configurable per interface hop penalty + + bugfixes + ======== + ++ batman-adv: Avoid uninitialized chaddr when handling DHCP ++ batman-adv: Fix own OGM check in aggregated OGMs + + + - - 2020.2 (2020-07-01) ++ 2020.3 (2020-08-25) + =================== + - * support latest kernels (4.4 - 5.8) ++ * support latest kernels (4.4 - 5.9) + * coding style cleanups and refactoring - * dropped support for kernels < 4.4 - * re-enabled link speed detection for interfaces without auto negotiation ++ * introduce a configurable per interface hop penalty ++ * bugs squashed: ++ ++ - avoid uninitialized chaddr when handling DHCP ++ - fix own OGMv2 check in aggregation receive handling + +batctl +------ + +:: + + $ git describe origin/master - v2020.1-8-g51b0746 - $ range=v2020.1..v2020.1-8-g51b0746 ++ v2020.2-3-g2c893e3 ++ $ range=v2020.2..v2020.2-3-g2c893e3 + $ git shortlog --email --no-merges "${range}" + $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq + + + various + ======= + - batctl: Disable automatic destruction of empty meshifs - - coding style cleanup/refactoring - ================================ - - batctl: Change/fix a throughput override function name - - bugfix - ====== - - batctl: fix endianness when reading radiotap header - batctl: Only remove batadv interface on hardif reduction ++ batctl: Add per interface hop penalty command + + - 2020.2 (2020-07-01) ++ 2020.3 (2020-08-25) + =================== + - * coding style cleanups and refactoring - * drop support for automatic destruction of empty meshifs - * bugs squashed: - - - Fix parsing of radiotap headers on big endian systems ++ * add per interface hop penalty command + +alfred +------ + +:: + + $ git describe origin/master - v2020.1-3-g112788d - $ range=v2020.1..v2020.1-3-g112788d ++ v2020.2-2-g921940b ++ $ range=v2020.2..v2020.2-2-g921940b + $ git shortlog --email --no-merges "${range}" + $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq + ++ Sven Eckelmann <sven(a)narfation.org> (1): ++ batctl: Sync batman-adv netlink uapi header + - Sven Eckelmann <sven(a)narfation.org> (2): - alfred: Switch mailing list subscription page - alfred: Use primary/secondary to describe server roles - - 2020.2 (2020-07-01) ++ 2020.3 (2020-08-25) + =================== + - * Rephrase names of server roles ++ * synchronization of batman-adv netlink header + +Happy routing, + +The B.A.T.M.A.N. team diff --cc open-mesh/WikiStart.rst index c105d27,0000000..df6b2f0 mode 100644,000000..100644 --- a/open-mesh/WikiStart.rst +++ b/open-mesh/WikiStart.rst @@@ -1,60 -1,0 +1,53 @@@ +.. SPDX-License-Identifier: GPL-2.0 + +==================== +Welcome to Open Mesh +==================== + +**This page serves as development platform for a collection of tools to +build free and open mesh networks.** + +B.A.T.M.A.N. +============ + +B.A.T.M.A.N. (better approach to mobile ad-hoc networking) is a routing +protocol for multi-hop ad-hoc mesh networks. This is the main +development website, we have source code, binary packages, documentation +and further information available: + +News +---- + +{{recent\_news}} + +- :doc:`news archive <News-archive>` + +Getting Started +--------------- + +- :doc:`Download <Download>` - Get our latest releases, get access to our + Git repositories +- :doc:`How to contribute <Contribute>` - How to report bugs, send + comments, patches or suggestions + +Documentation +------------- + +- :doc:`Branches explained <BranchesExplained>` - learn about the + B.A.T.M.A.N. flavours +- :doc:`B.A.T.M.A.N. Advanced Documentation </batman-adv/index>` +- :doc:`B.A.T.M.A.N. Daemon Documentation </batmand/index>` +- :doc:`A.L.F.R.E.D. </alfred/index>` - Almighty Lightweight Fact Remote Exchange + Daemon, can be used to distribute arbitrary information over a mesh + network +- :doc:`FAQ <FAQ>` - Frequently Asked Questions +- :doc:`Experience <Experience>` - Collection of links to B.A.T.M.A.N. + press reviews, papers and projects using B.A.T.M.A.N. +- :doc:`Mailing List <MailingList>`/:doc:`IRC <IRC>` - contact us + +Merchandise +----------- + +- :doc:`Logos <Logos>` - download logos +- `T-Shirts <https://www.freewear.org/?org=B.A.T.M.A.N%2e>`__ - get + T-Shirts - - Other Projects on open-mesh.org - ------------------------------- - - - :doc:`FFTrace <FFTrace>` - freifunk traceroute tool - - :doc:`MeshLinux <MeshLinux>` - Linux distribution geared to build mesh - supernodes diff --cc open-mesh/random.rst index 92cc2b7,0000000..89005f6 mode 100644,000000..100644 --- a/open-mesh/random.rst +++ b/open-mesh/random.rst @@@ -1,40 -1,0 +1,30 @@@ +.. SPDX-License-Identifier: GPL-2.0 + +====== +Random +====== + +.. toctree:: + :maxdepth: 2 + :caption: Contents: + + + BATMANConcept + Emulation_Environment - FFTrace + Joker - MeshLinux + Logos + BatmanAdvBootLogo - NetworkingTechnologies + New_papers + News-archive + News-draft + ResearchProposals + Routing_scenarios + The-olsr-story + UsingBatmanGit - Gsoc2010-ideas - Gsoc2011-ideas - Gsoc2011-roadmap - Gsoc2012-ideas - Wbmv4-batmandev-agenda - Wbmv5-batmandev-agenda - Wbmv6-batmandev-agenda + +.. only:: subproject + + Indices + ======= + + * :ref:`genindex`

1 0

[doc] backup-redmine/2020-08-16,master: doc: open-mesh/Random (3715905)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit 37159056f230f31e1b5271ee082f5d45e44ebc93 Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 23:23:13 2020 +0000 doc: open-mesh/Random >--------------------------------------------------------------- 37159056f230f31e1b5271ee082f5d45e44ebc93 open-mesh/Random.textile | 5 ----- 1 file changed, 5 deletions(-) diff --git a/open-mesh/Random.textile b/open-mesh/Random.textile index 6a57c02..1ab694c 100644 --- a/open-mesh/Random.textile +++ b/open-mesh/Random.textile @@ -2,8 +2,3 @@ h2. Merchandise * [[Logos|Logos]] - download logos * "T-Shirts":https://www.freewear.org/?org=B.A.T.M.A.N%2e - get T-Shirts - -h2. Other Projects on open-mesh.org - -* [[FFTrace|FFTrace]] - freifunk traceroute tool -* [[MeshLinux|MeshLinux]] - Linux distribution geared to build mesh supernodes \ No newline at end of file

1 0

[doc] backup-redmine/2020-08-16,master: doc: open-mesh/News-draft: 2020.3 stub (49bde59)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit 49bde592991c3d64bcb6d2577e05c87c73247c7f Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 13:51:19 2020 +0000 doc: open-mesh/News-draft: 2020.3 stub >--------------------------------------------------------------- 49bde592991c3d64bcb6d2577e05c87c73247c7f open-mesh/News-draft.textile | 93 ++++++++++++++------------------------------ 1 file changed, 30 insertions(+), 63 deletions(-) diff --git a/open-mesh/News-draft.textile b/open-mesh/News-draft.textile index 3339aa5..263a2e2 100644 --- a/open-mesh/News-draft.textile +++ b/open-mesh/News-draft.textile @@ -1,78 +1,62 @@ -h1. DRAFT: Batman-adv 2020.2 released +h1. DRAFT: Batman-adv 2020.3 released -Jul 1th, 2020. Today the B.A.T.M.A.N. team publishes the July 2020 update to batman-adv, batctl and alfred! This release drops the support for the end-of-life Linux kernel versions (<4.3). The B.A.T.M.A.N. V throughput detection for ethernet devices without auto-negotiation was enabled again to support more configurations out of the box. Most users will still require a manual setting via the throughput_override for such devices. The automatic removal of (empty) batadv mesh interfaces during a <code>batctl meshif bat0 interface del ....</code> was dropped. An explicit <code>batctl meshif bat0 interface destroy</code> is now required to remove the batadv interface. Also several bugfixes and code cleanups are included in this version. +Aug 25th, 2020. Today the B.A.T.M.A.N. team publishes the August 2020 update to batman-adv, batctl and alfred! An additional hop penalty can now be configured on a per interface basis. Also several bugfixes and code cleanups are included in this version. As the kernel module always depends on the Linux kernel it is compiled against, it does not make sense to provide binaries on our website. As usual, you will find the signed tarballs in our download section: -https://downloads.open-mesh.org/batman/releases/batman-adv-2020.2/ +https://downloads.open-mesh.org/batman/releases/batman-adv-2020.3/ h2. Thanks Thanks to all people sending in patches: -* Al Viro <viro(a)zeniv.linux.org.uk> -* Antonio Quartulli <a(a)unstable.cc> -* Cong Wang <xiyou.wangcong(a)gmail.com> -* Marek Lindner <mareklindner(a)neomailbox.ch> * Linus L��ssing <linus.luessing(a)c0d3.blue> * Simon Wunderlich <sw(a)simonwunderlich.de> * Sven Eckelmann <sven(a)narfation.org> -* YueHaibing <yuehaibing(a)huawei.com> and to all those that supported us with good advice or rigorous testing: -* Andreas Ziegler <az(a)freifunk-altdorf.de> -* Matthias Schiffer <mschiffer(a)universe-factory.net> -* Taehee Yoo <ap420073(a)gmail.com> +* Antonio Quartulli <a(a)unstable.cc> h2. batman-adv <pre> $ git describe origin/master -v2020.1-13-ga042622f -$ range=v2020.1..v2020.1-13-ga042622f +v2020.2-7-geded19e9 +$ range=v2020.2..v2020.2-7-geded19e9 $ git shortlog --email --no-merges "${range}" $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq -new kernel support -================== - - batman-adv: Drop support for Linux < 4.4 - coding style cleanup/refactoring ================================ - batman-adv: get rid of pointless access_ok() - batman-adv: use rcu_replace_pointer() where appropriate - batman-adv: partially revert dynamic lockdep key changes - batman-adv: change addr_list_lock back to static key - batman-adv: Fix spelling error in term buffer - batman-adv: trace: Drop unneeded types.h include - batman-adv: Utilize prandom_u32_max for random [0, max) values - batman-adv: Sync config options for external module builds - batman-adv: Use alphabetical order for external build config options + batman-adv: Switch mailing list subscription page + batman-adv: Fix typos and grammar in documentation -BATMAN V -========= +various +======= - batman-adv: Revert "disable ethtool link speed detection when auto negotiation off" + batman-adv: Introduce a configurable per interface hop penalty bugfixes ======== + batman-adv: Avoid uninitialized chaddr when handling DHCP + batman-adv: Fix own OGM check in aggregated OGMs - -2020.2 (2020-07-01) +2020.3 (2020-08-25) =================== -* support latest kernels (4.4 - 5.8) +* support latest kernels (4.4 - 5.9) * coding style cleanups and refactoring -* dropped support for kernels < 4.4 -* re-enabled link speed detection for interfaces without auto negotiation +* introduce a configurable per interface hop penalty +* bugs squashed: + - avoid uninitialized chaddr when handling DHCP + - fix own OGMv2 check in aggregation receive handling </pre> @@ -80,8 +64,8 @@ h2. batctl <pre> $ git describe origin/master -v2020.1-8-g51b0746 -$ range=v2020.1..v2020.1-8-g51b0746 +v2020.2-3-g2c893e3 +$ range=v2020.2..v2020.2-3-g2c893e3 $ git shortlog --email --no-merges "${range}" $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq @@ -89,28 +73,13 @@ $ git log --no-merges "${range}"|grep -e '\(Reported\|Tested\|Acked\|Reviewed-by various ======= - batctl: Disable automatic destruction of empty meshifs - -coding style cleanup/refactoring -================================ - - batctl: Change/fix a throughput override function name - -bugfix -====== - - batctl: fix endianness when reading radiotap header - batctl: Only remove batadv interface on hardif reduction + batctl: Add per interface hop penalty command -2020.2 (2020-07-01) +2020.3 (2020-08-25) =================== -* coding style cleanups and refactoring -* drop support for automatic destruction of empty meshifs -* bugs squashed: - - - Fix parsing of radiotap headers on big endian systems +* add per interface hop penalty command </pre> @@ -118,20 +87,18 @@ h2. alfred <pre> $ git describe origin/master -v2020.1-3-g112788d -$ range=v2020.1..v2020.1-3-g112788d +v2020.2-2-g921940b +$ range=v2020.2..v2020.2-2-g921940b $ git shortlog --email --no-merges "${range}" $ git log --no-merges "${range}"|grep -e '$Reported\|Tested\|Acked\|Reviewed-by\|Co-authored-by$-by'|sed 's/.*:/*/'|sort|uniq +Sven Eckelmann <sven(a)narfation.org> (1): + batctl: Sync batman-adv netlink uapi header -Sven Eckelmann <sven(a)narfation.org> (2): - alfred: Switch mailing list subscription page - alfred: Use primary/secondary to describe server roles - -2020.2 (2020-07-01) +2020.3 (2020-08-25) =================== -* Rephrase names of server roles +* synchronization of batman-adv netlink header </pre>

1 0

[doc] backup-redmine/2020-08-16,master: doc: Drop outdated pages (6921383)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit 692138343ec89f5fba98c7bfdd9424b41731a42a Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 23:40:17 2020 +0200 doc: Drop outdated pages >--------------------------------------------------------------- 692138343ec89f5fba98c7bfdd9424b41731a42a ...Batman-adv-openwrt-config-sophisticated.textile | 208 --------------------- batman-adv/Batman-adv-openwrt-config.textile | 53 +----- batman-adv/ELP.textile | 3 +- .../Multi-link-optimizations-technical.textile | 80 -------- batman-adv/Multi-link-optimize-old.textile | 39 ---- .../Network-wide-multi-link-optimization.textile | 2 +- batman-adv/interface_detection.dia | Bin 1723 -> 0 bytes batman-adv/interface_detection.dia.jpg | Bin 22506 -> 0 bytes batman-adv/primary_flag.dia | Bin 1548 -> 0 bytes batman-adv/primary_flag.png | Bin 16010 -> 0 bytes open-mesh/FFTrace.textile | 11 -- open-mesh/MeshLinux.textile | 7 - open-mesh/NetworkingTechnologies.textile | 33 ---- 13 files changed, 3 insertions(+), 433 deletions(-) diff --git a/batman-adv/Batman-adv-openwrt-config-sophisticated.textile b/batman-adv/Batman-adv-openwrt-config-sophisticated.textile deleted file mode 100644 index 65dd874..0000000 --- a/batman-adv/Batman-adv-openwrt-config-sophisticated.textile +++ /dev/null @@ -1,208 +0,0 @@ -{{include(open-mesh:TemplateDeprecated)}} - -h1. Example B.A.T.M.A.N. configuration snippets on OpenWRT - -*Note*: The script below was written for batman-adv 2012.4.0 and older. With batman-adv 2013.0.0 came a new syntax as explained [[Batman-adv-openwrt-config|here]]. - -h2. 3-radio mesh-node, with one open VAP for roaming clients access, and another with WPA - -You should run the script inside the OpenWRT host, assuming: -* batman-adv is compiled in, or installed (opkg install kmod-batman-adv). -* your uci config resembles a vanilla OpenWRT (no esoteric setups done yet) -* all radios are connected -* drivers for all radios are compiled in or installed (ex. opkg install kmod-ath9k-htc) -* all your radios/drivers support adhoc mode properly (unfortunately most USB don't, as of May 2012) -* main radio (radio0) supports multiple, mixed VAPs (adhoc and ap) - -You can use this configuration (run this script) on a node with just 1 (or 2) radio(s) too, but you'll experience bandwidth degradation when passing through that point. - -<pre> -#!/bin/sh - -### Main radio0 will broadcast one AP with no encryption, another AP with WPA2, -### and both interfaces will be bridged together with eth0 and bat0 -### Another VAP in adhoc mode is added to main radio0, -### as well as adhoc networks in radio1 and radio2 if they are present. -### All three adhoc networks are added to bat0 and thus managed by batman-adv - -### Node-specific settings -export HOSTNAME="meshnodeX" -export IP="10.x.x.x" -export WPA_ESSID="$HOSTNAME.wpa" -export WPA_KEY="password" - -### These parameters should be consistent across all nodes -export NETMASK="255.0.0.0" -export DNS="" -export GATEWAY="" -export PUBLIC_ESSID="3radio.mesh" -export MESH0_BSSID="CA:CA:CA:CA:CA:00" -export MESH0_ESSID="mesh0" -export MESH0_CHANNEL="1" -export MESH1_MODE="adhoc" -export MESH1_BSSID="CA:CA:CA:CA:CA:01" -export MESH1_ESSID="mesh1" -export MESH1_CHANNEL="11" -export MESH2_MODE="adhoc" -export MESH2_BSSID="CA:CA:CA:CA:CA:02" -export MESH2_ESSID="mesh2" -export MESH2_CHANNEL="6" - -### Ensure of populating /etc/config/wireless with -### autodetected wifi-device entries (radioX) -### to get all list_capab and hwmode correct. Otherwise -### OpenWRT might fail to configure the radio properly. -wifi detect >>/etc/config/wireless - -### Clear preexisting wifi-iface sections to avoid conflicts or dups -( for i in `seq 0 9` ; do echo "delete wireless.@wifi-iface[]" ; done ) | uci batch -q - -### Create /etc/config/batman-adv if it's not there yet. -uci import -m batman-adv </dev/null - -echo " -set system.(a)system[0].hostname=$HOSTNAME - -set batman-adv.bat0=mesh -set batman-adv.bat0.interfaces='mesh0 mesh1 mesh2' - -set network.lan.ipaddr=$IP -set network.lan.netmask=$NETMASK -set network.lan.dns='$DNS' -set network.lan.gateway=$GATEWAY -set network.lan.ifname='eth0 bat0' -set network.bat0=interface -set network.bat0.ifname=bat0 -set network.bat0.proto=none -set network.bat0.mtu=1500 -set network.mesh0=interface -set network.mesh0.proto=none -set network.mesh0.mtu=1528 -set network.mesh1=interface -set network.mesh1.proto=none -set network.mesh1.mtu=1528 -set network.mesh2=interface -set network.mesh2.proto=none -set network.mesh2.mtu=1528 - -set wireless.radio0=wifi-device -set wireless.radio0.channel=$MESH0_CHANNEL -set wireless.radio0.disabled=0 -set wireless.radio0.phy=phy0 -set wireless.radio0.macaddr= - -set wireless.radio1=wifi-device -set wireless.radio1.channel=$MESH1_CHANNEL -set wireless.radio1.disabled=0 -set wireless.radio1.phy=phy1 -set wireless.radio1.macaddr= - -set wireless.radio2=wifi-device -set wireless.radio2.channel=$MESH2_CHANNEL -set wireless.radio2.disabled=0 -set wireless.radio2.phy=phy2 -set wireless.radio2.macaddr= - -add wireless wifi-iface -set wireless.(a)wifi-iface[-1].device=radio0 -set wireless.(a)wifi-iface[-1].encryption=none -set wireless.(a)wifi-iface[-1].network=lan -set wireless.(a)wifi-iface[-1].mode=ap -set wireless.(a)wifi-iface[-1].ssid='$PUBLIC_ESSID' - -add wireless wifi-iface -set wireless.(a)wifi-iface[-1].device=radio0 -set wireless.(a)wifi-iface[-1].encryption=psk2 -set wireless.(a)wifi-iface[-1].key='$WPA_KEY' -set wireless.(a)wifi-iface[-1].network=lan -set wireless.(a)wifi-iface[-1].mode=ap -set wireless.(a)wifi-iface[-1].ssid='$WPA_ESSID' - -add wireless wifi-iface -set wireless.(a)wifi-iface[-1].device=radio0 -set wireless.(a)wifi-iface[-1].encryption=none -set wireless.(a)wifi-iface[-1].network=mesh0 -set wireless.(a)wifi-iface[-1].mode=adhoc -set wireless.(a)wifi-iface[-1].bssid=$MESH0_BSSID -set wireless.(a)wifi-iface[-1].ssid='$MESH0_ESSID' -set wireless.(a)wifi-iface[-1].mcast_rate=11000 - -add wireless wifi-iface -set wireless.(a)wifi-iface[-1].device=radio1 -set wireless.(a)wifi-iface[-1].encryption=none -set wireless.(a)wifi-iface[-1].network=mesh1 -set wireless.(a)wifi-iface[-1].mode=$MESH1_MODE -set wireless.(a)wifi-iface[-1].bssid=$MESH1_BSSID -set wireless.(a)wifi-iface[-1].ssid='$MESH1_ESSID' -set wireless.(a)wifi-iface[-1].mcast_rate=11000 - -add wireless wifi-iface -set wireless.(a)wifi-iface[-1].device=radio2 -set wireless.(a)wifi-iface[-1].encryption=none -set wireless.(a)wifi-iface[-1].network=mesh2 -set wireless.(a)wifi-iface[-1].mode=$MESH2_MODE -set wireless.(a)wifi-iface[-1].bssid=$MESH2_BSSID -set wireless.(a)wifi-iface[-1].ssid='$MESH2_ESSID' -set wireless.(a)wifi-iface[-1].mcast_rate=11000 -commit" \ -| uci batch - -</pre> - -h3. What if radio1 and/or radio2 don't support adhoc mode properly? - -If the driver is buggy, unstable, or throughput is lower in adhoc mode, you can fallback to using an ap/client design, at the cost of redundancy. - -simply customize the MESHx_ parameters on a per-node basis: -<pre> -### Relative physical location: node1 <--ch6--> node2 <--ch11--> node3 - -# On node1 -export MESH1_MODE="client" -export MESH1_BSSID="" -export MESH1_ESSID="node2.ptp-channel6" -export MESH1_CHANNEL="6" - -# On node2 -export MESH1_MODE="ap" -export MESH1_BSSID="" -export MESH1_ESSID="node2.ptp-channel6" -export MESH1_CHANNEL="6" -export MESH2_MODE="client" -export MESH2_BSSID="" -export MESH2_ESSID="node3.ptp-channel11" -export MESH2_CHANNEL="11" - -# On node3 -export MESH1_MODE="ap" -export MESH1_BSSID="" -export MESH1_ESSID="node3.ptp-channel11" -export MESH1_CHANNEL="11" -</pre> - -h3. I like the script but I don't need/want an extra VAP with WPA2, how can I disable it? - -Just remove or comment out the corresponding section -<pre> -#add wireless wifi-iface -#set wireless.(a)wifi-iface[-1].device=radio0 -#set wireless.(a)wifi-iface[-1].encryption=psk2 -#set wireless.(a)wifi-iface[-1].key='$WPA_KEY' -#set wireless.(a)wifi-iface[-1].network=lan -#set wireless.(a)wifi-iface[-1].mode=ap -#set wireless.(a)wifi-iface[-1].ssid='$WPA_ESSID' -</pre> - -h3. What if I am running batman-adv 2013.0.0 or later ? - -Modify the bat0 and mesh0 networks to be as follows -<pre> -set network.bat0=interface -set network.bat0.ifname=bat0 -set network.bat0.proto=none -set network.bat0.mtu=1528 -set network.mesh0=interface -set network.mesh0.proto=batadv -set network.mesh0.mtu=1528 -set network.mesh0.mesh=bat0 -</pre> \ No newline at end of file diff --git a/batman-adv/Batman-adv-openwrt-config.textile b/batman-adv/Batman-adv-openwrt-config.textile index 9e935b5..9cf3b72 100644 --- a/batman-adv/Batman-adv-openwrt-config.textile +++ b/batman-adv/Batman-adv-openwrt-config.textile @@ -124,9 +124,7 @@ config interface 'my_bat_vlan1' Any VLAN specific option can still be listed in the batadv meshif configuration for 'bat0' and in that case they will affect the behaviour only for the plain mesh interface (i.e. bat0 - the untagged LAN). -h2. Batman-adv 2013.0.0 till 2019.0-2 - -With batman-adv 2013.0.0 the OpenWrt package was converted to integrate with OpenWrt's netifd system. This has some impact on the configuration but not much. +h2. Batman-adv 2019.0-2 and older The wireless configuration in '/etc/config/wireless': <pre> @@ -199,52 +197,3 @@ config interface 'bat0_hardif_wlan' option mesh 'bat0' option routing_algo 'BATMAN_V' </pre> \ No newline at end of file - -h2. Batman-adv 2012.4.0 and older - -The wifi interface needs to be configured in '/etc/config/wireless': -<pre> -config wifi-device 'radio0' - option [..] - -config wifi-iface 'wmesh' - option device 'radio0' - option ifname 'adhoc0' - option network 'bat0_hardif_wlan' - option mode 'adhoc' - option ssid 'mesh' - option 'mcast_rate' '18000' - option bssid '02:CA:FE:CA:CA:40' -</pre> - -It is assumed you configured the 'wifi-device' depending on your requirements and your hardware. The interesting part is the 'wifi-iface' stanza with its options: - -* 'device' points back to your radio (wifi-device) interface -* 'ifname' allows you to specify an arbitrary name for your adhoc interface (which we are going to re-use later) -* 'network' points to the corresponding stanza in '/etc/config/network' -* 'mode' defines the wifi mode (adhoc in our case) -* 'mcast_rate' helps to avoid low bandwidth routes (routes with a lower throughput rate than the mcast rate will not be visible to batman-adv) -* 'ssid' and 'bssid' are basic wireless settings you might want to set to your liking - -More information can be found in the "OpenWrt wireless configuration":https://wiki.openwrt.org/doc/uci/wireless - -Configure the MTU in '/etc/config/network' -<pre> -config interface 'bat0_hardif_wlan' - option ifname 'adhoc0' - option mtu '1528' - option proto 'none' -</pre> - -The stanza's name 'mesh' as well as the ifname option have to match your wireless configuration. - -Batman-adv is configured through its own configuration in '/etc/config/batman-adv': -<pre> -config mesh 'bat0' - option interfaces 'adhoc0' - option 'aggregated_ogms' - option 'ap_isolation' - option [..] -</pre> - -The 'interfaces' option is the key element here, as it tells batman-adv which interface(s) to run on. All the other options enable / disable / tweak all the batman-adv features you can also access at runtime through batctl. \ No newline at end of file diff --git a/batman-adv/ELP.textile b/batman-adv/ELP.textile index 66b1c94..c4db231 100644 --- a/batman-adv/ELP.textile +++ b/batman-adv/ELP.textile @@ -1,8 +1,7 @@ -h1. This page is OUTDATED! +{{include(open-mesh:TemplateDeprecated)}} This page is not in sync with the actual implementation. Neither the packet format matches nor the use as unicast probe frames is explained. - h1. Echo Location Protocol (ELP) !https://upload.wikimedia.org/wikipedia/commons/e/e1/Animal_echolocation.svg!:https://en.wikipedia.org/wiki/File:Animal_echolocation.svg diff --git a/batman-adv/Multi-link-optimizations-technical.textile b/batman-adv/Multi-link-optimizations-technical.textile deleted file mode 100644 index ee7cc48..0000000 --- a/batman-adv/Multi-link-optimizations-technical.textile +++ /dev/null @@ -1,80 +0,0 @@ -h1. This page is OUTDATED! - -This page was superseded with an update for the feature, [[Network-wide-multi-link-optimization|documentered here]] - -h1. Multi-link Optimizations (technical documentation) - -Mesh nodes with multiple physical links to other mesh nodes can use this capability to gain various benefits, for example: - -* bonding: use multiple links for the same data stream to boost bandwidth -* interface alternating: send on other interfaces (and therefore other channels, if configured correctly) to minimize multi-hop bandwidth degradation - -These features operate under the following premises: - -* multiple links exist between two neighbors -* the routing decision is independent of the link a packet is received on - -Notably, these multi-interface features are independent of B.A.T.M.A.N. and could work with any routing algorithm/metric. The routing software only needs to detect that multiple peers belong to the same neighbor mesh node. This is also not related to Multi-Path Routing, which would require multiple independent links. - -The multi-link optimization procedure can be broken into these steps: - -# detect that a neighbor is reachable via multiple links -# select suitable candidates among the the available links -# use these multiple links for various manipulations - -h2. Link detection - -The detection part is specific to batman-adv. To understand how the link detection works it is necessary to recall how batman-adv handles multiple interfaces: - -The MAC address (Originator address) of the primary interface (usually the first added interface) is known throughout the mesh as OGMs containing the MAC address are broadcasted on all available interfaces with a big TTL. This ensures mesh nodes only reachable via secondary interfaces will know about this node. All remaining interfaces are "secondary" interfaces which are only known in the single hop neighborhood. The OGMs containing their MAC addresses are broadcasted with a small TTL on the respective interface only. These secondary OGMs uniquely serve the purpose of detecting local neighbors and estimate the link quality towards them. - -+Note+: It does not matter which interface is the primary interface, it is only used to have a single Originator known in the mesh, thus avoiding OGM overhead. - -To detect whether or not a neighbor has multiple links towards a node it suffices to analyze the incoming primary OGM. If the primary OGM is received via multiple interfaces a node can (almost) safely assume having multiple links towards its Originator. However, the OGM protocol demands to forward all incoming OGMs for the path detection which could confuse the multi-link detection because it has to distinguish single hop neighbors from neighbors that are further away. To resolve the issue the PRIMARIES_FIRST_HOP flag was introduced. Only primary OGMs are allowed to carry it and every node removes the flag before rebroadcasting the OGM. - -!primary_flag.png! - -In the illustration, A receives the OGM via A1 (sent out by B1) and A2 (sent out by B2) - and detects that these two Links A1-B1, A2-B2 exist. These link candidates are stored and further filtered - -h2. Candidate selection - -To be usable for the target features, the links must have a similar quality and should operate on a different frequency. Therefore, in batman-adv only links are further considered if: - -* the link quality is not much worse than the best link within a certain TQ threshold (now: 50) -* the link destination is different from the other already selected candidates - e.g. if A1-B1 was selected, A2-B1 would be forbidden as they would operate on the same frequency and therefore would interfere with each other - -In the illustration below, A1-B1 and A2-B2 meet this criteria and are therefore selected as candidates. - -!interface_detection.dia.jpg! - -Implementation note: Detected links are internally stored within the originator structure of the primary originator. The candidates are internally called "bonding candidates" - -h2. Using the candidates - -Once a neighbor mesh node has been selected by the classic routing algorithm, the individual links can now be equally be selected to send data to the next neighbor without tampering with routing. Possible mechanisms exploiting this fact are described below. - -More Info can be found in the [[multi-link-optimize|Multi-Link Optimization]] feature description page. - -Implementation note: the manipulations are hooked in the tx path when selecting a neighbor in find_router() - -h3. Interface Alternating - -Interface Alternating basically tries to forward a packet on another interface than on which the frame was received. This is only done if suitable candidate with another source interface exists. - -!alternation_chain.dia.jpg! - -This feature is always enabled. - -h3. Bonding - -In bonding, we split a single stream and distribute the packets over multiple links. Packets are sent in a round-robin fashion over the available candidates. This can be used to increase the throughput for data streams from node A to node B. In a perfect environment the throughput is n times higher, where n is the number of candidates. - -!bonding_roundrobin.dia.jpg! - -Practice shows however that the theoretical bandwidth gain is not reached - even if the links are physically disjoint. E.g. using a 2.4 GHz and a 5 GHz with a modulation of 54 Mbit/s both has shown to gain 150% bandwith boost over a single link, not 200% as expected in theory. Possible reasons may be the packet reordering and slightly different queue load caused by random backoffs, retries, etc. For these reasons, bonding must be enabled explicitly. - -h2. Discussion and further work - -As explained, the bonding mode stays behind the theoretical expectations. The packet loss based TQ metric is also not optimal for selecting bonding candidates - a bandwidth based metric would be more suitable for this task, as slow links will slow down the whole aggregated connection. - -These shortcomings would be subject to further improvements of this featureset. \ No newline at end of file diff --git a/batman-adv/Multi-link-optimize-old.textile b/batman-adv/Multi-link-optimize-old.textile deleted file mode 100644 index c5e1254..0000000 --- a/batman-adv/Multi-link-optimize-old.textile +++ /dev/null @@ -1,39 +0,0 @@ -h1. This page is OUTDATED! - -This page was superseded with an update for the feature, [[Multi-link-optimize|documentered here]] - -h1. Multi-link Optimizations - -batman-adv is able to work on different interfaces - multiple WiFi interfaces, but also Ethernet interfaces. If two or more nodes are interconnected through batman-adv on multiple interfaces, they may leverage some features to increase throughput, which are explained here. - -For more technical information, please see our [[multi-link-optimizations-technical|technical multi-link documentation]] - -h2. Interface alternating - -With Interface alternating, batman-adv forwards frames on a different interface than on which the frame was received. The purpose of this alternation is to reduce interference (we can either send or receive on a WiFi interface at one time) and to balance the network load better on the available interfaces, and eventually increases throughput. The mechanism is illustrated below for a chain of nodes with two interfaces. - -!alternation_chain.dia.jpg! - -Interface alternating is only performed if the two candidate links to the next hop have a similar quality. This feature is enabled by default and does not have to be enabled manually. - -h2. Interface bonding - -When multiple links (with similar transmit quality) are available to the same neighbor, batman-adv may distribute the frames to be sent over these available links. The individual frames are sent over the links in a round robin fashion as illustrated below. Using this technique, the throughput may be increased by the number of links involved in the bonding. In practical tests over two WiFi links, we have seen more than 50% of throughput gain. - -!bonding_roundrobin.dia.jpg! - -However, if the links have different speeds (batman-adv won't detect that), the throughput may even decrease due to the slower link slowing down the whole bonding. Therefore, this feature must be explicitly enabled. One may enable it for certain nodes in a known environment, the same setting does not have to be applied mesh wide. - -To enable bonding, use batctl: - -<pre> -batctl bonding enable -</pre> - -h2. Throughput gain - -At the [[open-mesh:2010-06-13-wbm2010-bracciano|WirelessBattleMesh v3]] we performed throughput tests to measure the gain of the various modes: - -!bonding_2hops.png! - -This graph shows a 3 node setup each with 2 802.11abg wifi interfaces connected to the mesh network (1x 2.4GHz and 1x 5.8GHz). As you can see the alternating mode is able to maintain the performance. The old default behavior sometimes manages to maintain a similar level of performance but is rather unstable. diff --git a/batman-adv/Network-wide-multi-link-optimization.textile b/batman-adv/Network-wide-multi-link-optimization.textile index b42b8e6..f70ca47 100644 --- a/batman-adv/Network-wide-multi-link-optimization.textile +++ b/batman-adv/Network-wide-multi-link-optimization.textile @@ -1,6 +1,6 @@ h1. Network Wide Multi Link Optimization (technical documentation) -The original batman-adv multi link optimization ([[Multi-link-optimizations-technical]]) works on a local level - which is fairly easy to implement and works well as long as as all nodes are configured the same way and have similar link qualities. But local decision may be suboptimal network wide - for example, consider some dual radio mesh nodes where the first node starts on 5 GHz, but after some hops (alternating between 2.4 and 5 GHz) it stops at a 2.4 GHz-only device. If it had chosen the other (2.4 Ghz) link in the first hop, choosing the same frequency twice could have been avoided, and the total path quality would be better. +The original batman-adv multi link optimization worked on a local level - which is fairly easy to implement and works well as long as as all nodes are configured the same way and have similar link qualities. But local decision may be suboptimal network wide - for example, consider some dual radio mesh nodes where the first node starts on 5 GHz, but after some hops (alternating between 2.4 and 5 GHz) it stops at a 2.4 GHz-only device. If it had chosen the other (2.4 Ghz) link in the first hop, choosing the same frequency twice could have been avoided, and the total path quality would be better. !alternating-limited-view.png! diff --git a/batman-adv/interface_detection.dia b/batman-adv/interface_detection.dia deleted file mode 100644 index 21b9955..0000000 Binary files a/batman-adv/interface_detection.dia and /dev/null differ diff --git a/batman-adv/interface_detection.dia.jpg b/batman-adv/interface_detection.dia.jpg deleted file mode 100644 index 015df6e..0000000 Binary files a/batman-adv/interface_detection.dia.jpg and /dev/null differ diff --git a/batman-adv/primary_flag.dia b/batman-adv/primary_flag.dia deleted file mode 100644 index 2422806..0000000 Binary files a/batman-adv/primary_flag.dia and /dev/null differ diff --git a/batman-adv/primary_flag.png b/batman-adv/primary_flag.png deleted file mode 100644 index 0c3e3f8..0000000 Binary files a/batman-adv/primary_flag.png and /dev/null differ diff --git a/open-mesh/FFTrace.textile b/open-mesh/FFTrace.textile deleted file mode 100644 index 42cc471..0000000 --- a/open-mesh/FFTrace.textile +++ /dev/null @@ -1,11 +0,0 @@ -h1. FFTrace freifunk-traceroute - -*"fftrace"* is a console-based olsr-route-link-quality monitor - -Problem description: in OLSR mesh networks routing is very dynamic. Fine tools as traceroute give only a one-time impression, and mtr does not cope with routechanges as often as they happen. Both aim to measure and show the round trip times of packets as quality indicator. A wireless network consists of many diffrent-quality HF-connections, some with severe packet loss and some with less. The link-quality implementation of olsrd introduced a value do describe the quality of these connections. A tool to monitor the current route and the link-quality of its connections was missing. - -*"fftrace"* aims to fill this gap. It lists the single nodes of a route and displays the quality of the connections in between. Finally it shows the needed effort to get a packet through the whole route. - -The program uses ttl-manipulatet icmp-echo-requests to find the intermediate nodes and reads the udp-messages used and generated by the olsr-deamons to gather present information about the link-quality and neighbor-link-quality of node-node-pairs in the watched route. Due to this structure there is no instant information on a new node-node-pair. Until the information is available it is left blank and unregarded by the whole-route-etx. The code tries to be modular enough to exchange parts if desired, such as the udp-part by an olsr-plugin or the screen-output in something that can be diplayed graphical e.g. by olsr-viz. (gpl is gpl, be free to do so) - -* Download: https://downloads.open-mesh.org/fftrace/ diff --git a/open-mesh/MeshLinux.textile b/open-mesh/MeshLinux.textile deleted file mode 100644 index c4901a1..0000000 --- a/open-mesh/MeshLinux.textile +++ /dev/null @@ -1,7 +0,0 @@ -h1. MeshLinux - -Meshlinux is outdated and really old. The project hasn't been updated in a long time. It is unlikely that anyone will take up development on it again. We recommend to use OpenWrt for X86 platform. The only purpose we can think of today is software archaeology or the use of very old legacy WiFi hardware. - -Nothing is as outdated as old software. Meshlinux was created in the early days of mesh networking at the Freifunk community in 2004. It is a CD-Image that can be used to convert old PCs into wireless supernodes. There is a old Meshlinux version based on Linux (Kernel 2.4.26) available at (https://wiki.c-base.org/coredump/MeshLinux) This old version is based on Slackware and not maintained anymore. However it works nicely with old Atmel USB wireless devices (which is a problem nowadays with recent kernels), Prism-based Chipsets and Orinoco. But you won't find any modern mesh routing protocol implementations in it. If you want to build a meshrouter with nothing better than a 486 CPU and this kind of old wireless devices because you have them lying around and want to do something useful with it, you may still use it and install a recent version of Layer 3 B.A.T.M.A.N. - you can download a statically compiled version from this server. You can not use Batman-advanced, since this would r equire a recent Linux kernel. Some userland mesh protocol daemons like OLSR might still work with it. - -Sooner or later this information will be removed. \ No newline at end of file diff --git a/open-mesh/NetworkingTechnologies.textile b/open-mesh/NetworkingTechnologies.textile deleted file mode 100644 index b7155ee..0000000 --- a/open-mesh/NetworkingTechnologies.textile +++ /dev/null @@ -1,33 +0,0 @@ -{{include(open-mesh:TemplateDeprecated)}} - -h1. Networking Technologies - -h2. Introduction. - -Batmand and batman advanced rely on underlining network transport technologies like WIFI for instance. -This page explains the interaction between them and batmand or batmand-adv. - -h2. Batmand and IBSS (also known as ad-hoc wifi networks) - -IBSS permit the connection of more than two devices together in a big IBSS network. -Let's say you have 3 devices A,B and C connected that way: A<->B<->C with: -* C being out of the range of A -* A and B being in range -* B and C being in range -* A,B,C having the same SSID (it's called BSSID in IBSS networks) - -Batmand will simply route the packets from A to C. - -Given the nature of IBSS networks, B doesn't need to have 2 wifi cards(one is enough), however A,B,C need to have the same BSSID. - -h2. Batman-adv - -Batman-adv can run on top of all ethernet-compatible networks, that means: -* wifi (station, ibss, master) -* ethernet -* bridges -* layer 2 vpns -* etc... - - -

1 0

[doc] backup-redmine/2020-08-16,master: doc: Drop old Wireless Battlemesh Agendas (b596c02)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit b596c02f02b890b1e508431c43a1e470f9adf2a5 Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 23:47:35 2020 +0200 doc: Drop old Wireless Battlemesh Agendas >--------------------------------------------------------------- b596c02f02b890b1e508431c43a1e470f9adf2a5 open-mesh/Wbmv4-batmandev-agenda.textile | 33 ----------------------- open-mesh/Wbmv5-batmandev-agenda.textile | 23 ---------------- open-mesh/Wbmv6-batmandev-agenda.textile | 46 -------------------------------- 3 files changed, 102 deletions(-) diff --git a/open-mesh/Wbmv4-batmandev-agenda.textile b/open-mesh/Wbmv4-batmandev-agenda.textile deleted file mode 100644 index 2e39df9..0000000 --- a/open-mesh/Wbmv4-batmandev-agenda.textile +++ /dev/null @@ -1,33 +0,0 @@ - -h1. Wireless Battle Mesh v4 - Developer Meeting Agenda - - - -h2. Party!!! - -Celebrate batman-adv 2.6.38 Linux kernel integration - - -h2. Discussion Topics - -* How to proceed with NDP integration? Smaller patchset for LQ exchange first or including OGM protocol changes? -* How to name OGM protocol part? Just OGM protocol? How to name NDP packets, NDP messages? -* Explicit BATMAN header + unified TTL handling? -* Making fragmentation more generic, adding Fragmentation header/layer? -* Reviewing OGM protocol specification proposal -* reviewing the new papers related to B.A.T.M.A.N. (Simon brings printed versions) -* discuss switching to git -* Creating an extra devel mailinglist? - - -h2. Testing - -* NDP -* Multicast patches -* Hidden-Node-Bcast problem -* New HNA announcement mechanism - - -h2. Coding - -* new HNA code could still need a bit of coding, in particular the lock/unlock mechanism has to be reviewed \ No newline at end of file diff --git a/open-mesh/Wbmv5-batmandev-agenda.textile b/open-mesh/Wbmv5-batmandev-agenda.textile deleted file mode 100644 index ae8c8dd..0000000 --- a/open-mesh/Wbmv5-batmandev-agenda.textile +++ /dev/null @@ -1,23 +0,0 @@ -h1. Wbmv5-batmandev-agenda - -h2. Celebration! - -Find something to celebrate. - -h2. Testing - -* Catwoman (Throughput test) -* ELP Testing (Is the code ready? poke Marek) -* Roaming - -h2. Concept discussion - -* ELP / OGMv2 -* Bandwidth based routing - -h2. Others - -* GSOC 2012 ideas refinement -* Explaining the batman-adv release cycle - the journey of a patch through the branches -* How to make David happy: hands-on session -* Donations diff --git a/open-mesh/Wbmv6-batmandev-agenda.textile b/open-mesh/Wbmv6-batmandev-agenda.textile deleted file mode 100644 index 77b1737..0000000 --- a/open-mesh/Wbmv6-batmandev-agenda.textile +++ /dev/null @@ -1,46 +0,0 @@ -h1. Wbmv6-batmandev-agenda - -h2. Celebration! - -Find something to celebrate (something different from last year). - -h2. Would be good if we do - -* Big Compat Dump, consisting in integration of: -** TVLV -** FRAGv2 -** CRC32 in TT -** Improve packet format design to avoid __packed (?) -** bandwidth meter packet format integration (? not sure if this does really hurt since bw meter uses icmp) - -h2. Testing - -* Catwoman (Throughput test) - -h2. Concept discussion - -* B.A.T.M.A.N. V (which includes bandwidth aware ELP and OGMv2) -* OGMv3 (no-tt OGM) -* Multicast (evolution) support inclusion -* "Howto to optimise the Bandwidth based GW selection":https://lists.open-mesh.org/mailman3/hyperkitty/list/b.a.t.m.a.n@lists.open-mesh.org/message/5FJF5RSSYJVX4R3TLCIZ2A3Z2BTXENHH/ -* Multiple L3 border Gateways to the same network: how to choose the best one assuming each of them has a different (L3) cost to the destination -* Some notes were taken in the past about possible improvements to the GW feature. We may want to consider them: -<pre>- Multiple gw for multiple networks -Tag the GW so that you can choose the best among a single class and keep a -list of the best gws (one per class). -Each class can connect the mesh to a different network, therefore we would -want to communicate to several gw at the same time, not in mutual exclusion.</pre> -* mac80211 like DebugFS structure -* Fragmentation v2: advertise MTUs to avoid re-fragmentation? - -h2. Talk - -* Catwoman II (aka Fox) - -h2. Others - -* -GSOC 2013- -** -decide mentors availability- -** -think about possible projects- -* Maint release cycle (i.e. Stable versions) -* Compat code maintenance (split kernel-module and out-of-the-kernel package development?) \ No newline at end of file

1 0

[doc] backup-redmine/2020-08-16,master: doc: Drop old GSOC pages (20a0ea2)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit 20a0ea2cdf115f1bf8ec0f5a21469b449f50d911 Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 23:54:11 2020 +0200 doc: Drop old GSOC pages Signed-off-by: Sven Eckelmann <sven(a)narfation.org> >--------------------------------------------------------------- 20a0ea2cdf115f1bf8ec0f5a21469b449f50d911 open-mesh/Gsoc2010-ideas.textile | 91 -------------------------- open-mesh/Gsoc2011-ideas.textile | 131 ------------------------------------- open-mesh/Gsoc2011-roadmap.textile | 15 ----- open-mesh/Gsoc2012-ideas.textile | 9 --- 4 files changed, 246 deletions(-) diff --git a/open-mesh/Gsoc2010-ideas.textile b/open-mesh/Gsoc2010-ideas.textile deleted file mode 100644 index f27ed34..0000000 --- a/open-mesh/Gsoc2010-ideas.textile +++ /dev/null @@ -1,91 +0,0 @@ -h1. Gsoc2010 ideas - -h2. Requirements for students - -* Excited interest in Mesh networking technologies -* Routing and networking knowledge in general -* Programming language: C - -h2. Recommended and/or useful - -These things are not a prerequisite but might be very useful and/or have to be learned during the GSoC anyway. -* [[Linux kernel coding style]] -* Reading the documentation of the algorithms: https://git.open-mesh.org/?p=batman-adv-doc.git https://tools.ietf.org/html/draft-wunderlich-openmesh-manet-routing-00 (used in the 'old' batmand) -* Being able to use virtual machines for debugging (i.e. qemu) -* Being able to install a mesh network with [[OpenWRT]] and batman-adv -* Kernel coding experiences (especially with how to handle and avoid memory management/race conditions/deadlocks) -* monitoring packet flow (using tcpdump / wireshark + wireshark batman-adv dissector) - - -h2. Ideas - - - -h3. link layer fragmentation / compression - -*Brief description:* Introducing link-layer fragmentation and header compression to offer alternative packet overhead solutions. - -Due to the packet encapsulation performed on each and every payload packet going through the mesh a small overhead is introduced. This overhead (24 bytes at the moment) must be dealt with by the admins deploying the batman-adv mesh. The most common solution is to increase the maximum packet size on all mesh interfaces to maintain the standard MTU of 1500 bytes on all client devices. While this works in most cases it does not work for everyone. Some devices / drivers don't support or allow bigger packets. Alternatively, each client has to reduce its MTU to 1476 bytes which is often not manageable. This project shall address the issue by either compressing the payload packets (e.g. VJ compression) and/or implement a lightweight link layer fragmentation and/or develop another solution. - - -h3. forward error correction - -*Brief description:* Forward error correction to avoid retransmissions in the mesh network. - -Wireless networks rely on a lossy transport medium - even under optimal conditions packets get lost and have to be retransmitted. This comes at an even greater cost since the air is a shared medium (only one participant can transmit data at any given time) retransmissions have a severe impact on the network's performance. In addition to the packets which have to travel over several hops towards their destination, higher layer protocols (e.g. TCP) expect ACK packets to travel back over the chain of connected mesh nodes otherwise the transmission starts anew. This project aims to avoid retransmissions by implementing a packet forward correction for the batman-adv payload packets. A parity packet would be injected into the packet flow which allows the receiving batman-adv node to calculate & deliver the lost packet without further retransmissions. The parity packet rate can by dynamically adjusted depending on the link quality towards the final destination. - - -h3. B.A.T.M.A.N. protocol overhead reduction - -*Brief description:* Split the currently used OGM packets into two separate types to reduce the amount of packets flooded in the neighborhood. - -The traffic B.A.T.M.A.N. IV generates to discover the best path through the network is quite low compared to other protocols, but especially when B.A.T.M.A.N. has many single hop neighbors which rebroadcast each others OGMs we see room for improvements. The project shall optimize the flooding algorithm by splitting the originator message into two different message types. The OGM will remain but only be used to flood the TQ throughout the network. The new message type (a name needs to be found) will contain the link qualities of the single hop neighbors only. This message won't be rebroadcasted and just reaches the local neighborhood. These local message can be sent much more often than the global TQ messages and thus reduce the traffic [nearly just create a linear growth of traffic with more nodes in the local neighborhood instead of a squared amount]. - - -h3. B.A.T.M.A.N. protocol convergence speed - -*Brief description:* Implement a fast changing environment detection algorithm to let mobile nodes adapt to their new surroundings faster. - -Fast moving nodes always have the problem of adjusting their routing information in time. They can choose to send more routing information, so that their environment can adjust to them but stationary nodes won't do the same and increase the mobile node's adaption time greatly. However, when a B.A.T.M.A.N. node detects that its local environment changed quickly, it will enter the starvation mode. In this mode the node will actively try to confirm a working route as fast as possible by sending a "batman ping" to its new neighbors. Each B.A.T.M.A.N. neighbor will try to forward the message to its destination, once arrived there it will travel back. If the mobile node receives the reply it can change its route towards the new neighbor without waiting for normal OGM flooding as the route has been verified. The goal of this project is to implement the starving mode together with the "batman ping". - - -h3. Optimize multicast performance - -*Brief description:* Avoid broadcasting multicast traffic to nodes not belonging to a certain multicast group. - -Multicast packets would have a great advantage especially in wireless mesh networks: The number of recipients of multicast packets (i.e. for zeroconf service announcements or audio/video streams) is a lot higher then with unicast packets. batman-adv currently handles multicast packets the same way as broadcast packets, they simply get flooded through the mesh network. Instead, those packets should be flooded on a subgraph containing the nodes of a certain multicast group and other connecting nodes only. - - -h3. Dynamic OGM intervals - -*Brief description:* A batman-adv node shall select originator interval rates according to the dense and dynamics in its closer environment. - -Depending on the usage scenario, people can adjust the bandwidth being used for batman-adv's route finding algorithms. Usually people were advised to increase the originator interval if the mesh network is small but needs fast route refresh rates or to decrease it if the mesh network is mostly a static setup with a lot of nodes. It would be great if a batman-adv node could determine the dynamics of the mesh network it is currently participating in on its own so that this option would not have to be administrated from a person anymore. For instance in combination with the 'B.A.T.M.A.N. protocol overhead reduction' project, the local broadcasts could be automatically increased if there are not that many direct neighbors. - - -h3. More batctl ping/traceroute options - -*Brief description:* Add useful options for administrating a mesh network to batctl. - -batctl is the nicer frontend for configuration and debugging batman-adv. There are still some options that could be added like showing the MTU of hops with traceroute along a path to make it easier to spot MTU issues. Or outgoing interface selection to be able to manually probe the way and quality of an alternative path. Feature proposals that make the administration of a complex mesh network more easy are welcome. - - -h3. Live VIS in map - -*Brief description:* Building tools that visualize the dot output of the vis server with additional gps coordinates on a map. - -With every new technology, a bridge to non-technical people should be provided as well. batman-adv is being used in routers of every-day users, that do not have an insight in the B.A.T.M.A.N. routing protocol itself, nevertheless a good visualization can widely increase the acceptance of a new technology and get young people interested in it. batman-adv has a built in vis-server which produces a raw dot-file when activated. With the help of graphviz-tools, those dot-files can be rendered as graphs which are still more interesting for 'technical' people. It would be great to have a tool that maps the information provided by the dot-output and additional geo coordinates in Google-Maps or OpenStreetMap (!OpenLayers) in realtime. Then 'normal' people could find out and solve dead zones without technical support all on their own without having to use fancy command line tools. This feature would be useful for anyone administrating (parts of) a mesh network. - - -h3. Multiple interfaces per node support in Mesh3D - -*Brief description:* Adapting Mesh3D to handle the new visualization format features from current batman-adv. - -Mesh 3D is an application written in C and maintained by one of the batman-developers which is able draw a 3 dimensional graph from batman's vis output in dot-format. The latest additions in batman-adv's vis output that now features a differentiated visualization, in that interface connections between nodes are now being shown separately. This new format feature has not been ported to Mesh3D yet. Also a concept for visualizing overlapping links in Mesh3D would probably have to be planed (adding transparency to Mesh3D for instance). - - -h3. Multiple switch ports for redundancy - -*Brief description:* Allow multiple bridge uplinks to wired networks - -It is often desired to have multiple uplinks to a (wired) switched network where the B.A.T.M.A.N. protocol is not used. This may be a data center or a core network where multiple or redundant connections are needed. However if the Mesh network device and Ethernet device are bridged on multiple nodes, bridge loops are created. Traditional measures like (R)STP don't help in this situation as they may disable the (good) Ethernet Links. A solution should be developed where we can use multiple (redundant) uplinks to the same core switch network while effectively avoiding switch loops. \ No newline at end of file diff --git a/open-mesh/Gsoc2011-ideas.textile b/open-mesh/Gsoc2011-ideas.textile deleted file mode 100644 index 7d68a72..0000000 --- a/open-mesh/Gsoc2011-ideas.textile +++ /dev/null @@ -1,131 +0,0 @@ -h1. Google Summer of Code 2011 - Ideas - - - -h2. batman-adv - -h3. Categories + Usecases to improve - -* Increasing Robustness -* Improving Mobility / Convergence Speed -* Improving Indoor performance -* Improving [[VoIP]] group multicast performance -* Improving nomadic scenarios (circular + chain topology) - - -h3. Specific Tasks - -h4. Dynamic OGM+NDP intervals - -*Brief description:* A batman-adv node shall select originator interval rates according to the dense and dynamics in its closer environment. - -Depending on the usage scenario, people can adjust the bandwidth being used for batman-adv's route finding algorithms. Usually people were advised to increase the originator interval if the mesh network is small but needs fast route refresh rates or to decrease it if the mesh network is mostly a static setup with a lot of nodes. It would be great if a batman-adv node could determine the dynamics of the mesh network it is currently participating in on its own so that this option would not have to be administrated from a person anymore. For instance in combination with the 'B.A.T.M.A.N. protocol overhead reduction' project, the local broadcasts could be automatically increased if there are not that many direct neighbors. - - -h4. OGM/NDP Hidden Node Problem Avoidance - -*Brief description:* Implement a suitable solution against hidden node broadcast packet problem. - -Especially in indoor scenarios where corners and thick walls are involved, hidden nodes can be a severe problem. Usually activating RTS/CTS is a common way of solving this problem at least to some degree. However, RTS/CTS can only be applied for unicast packets. Therefore a node sending a lot of data packets to another one, even with RTS/CTS those packets will interfere with BATMAN's broadcast packets (e.g. NDP packets or OGMs). The effect is, that the transmit quality of a node sending NDP packets which does not see the data packet transmission will greatly decrease. - -A suitable solution shall be implemented. This might be based on time slots, multiple interfaces or channel switching and might be implemented within batman-adv or an extra module. See [[batman-adv:bcast-hidden-node|here]] for a more detailed description of the problem and a solution proposal. - - -h4. OGM/NDP Congestion Avoidance - -*Brief description:* Implement a congestion avoidance mechanism for batman-adv's routing control packets against congestion caused by high data transfer rates. - -In case of a mesh network congested due to e.g. a lot of UDP packets, the packet loss rate of batman's OGMs or NDP packets increases severely, as neither OGMs/NDP nor UDP have a congestion control mechanism. For one thing, lost OGMs lead to a high degradation of batman's convergence speed. For another, lost NDP packets lead to route flipping; in the worst case some of these chosen routes are not usable. batman-adv has the advantage of being able to control the whole packet, both data and control packets, within the mesh network. Therefore time spaces for OGMs/NDP packets could be allocated, as already described in "OGM/NDP Hidden Node Problem Avoidance" / "Or prioritized queues could be introduced. - - -h4. B.A.T.M.A.N. Routing Framework - -*Brief description:* Separate B.A.T.M.A.N. routing algorithm from routing infrastructure in batman-adv. - -Many features in batman-adv are routing algorithm independent: Link layer routing, link layer fragmentation, multicast optimizations, NDP, interface bonding. To allow other routing algorithms to use this already well tested infrastructure, the B.A.T.M.A.N. routing algorithm shall be separated more clearly from the routing infrastructure. Ultimately, other routing protocols shall be able to control the kernel space mesh routing infrastructure from userspace. - - -h4. Improve Throughput Bonding - -*Brief description:* Improve performance gain of the throughput bonding mode. - -The current bonding mode implementation simply sends a data packet in a round robin fashion to the interfaces available for bonding. However, the throughput gain seems to be only about 66% instead of doubling the throughput in case of TCP, even if there is no other interference involved. This seems to be the case due to a heavy reordering issue. -Furthermore, the interface with the lowest capacity is a bottleneck for the throughput bonding performance: In case of one interface of 1MBit/s throughput and five more with 2MBit/s throughput, the actual throughput will be just 6x 1MBit/s. However in mixed wireless and wired networks, it might be desirable to accumulate the throughput of a Gigabit-Ethernet interface and a 802.11g interface. -Therefore checks for links' capacities (detect full queues etc.) to gain performance of IF1 + IF2 + ... instead of min(IF1, IF2, ...) shall be implemented. If an interface is busy, it shall be skipped in this round robin cycle. Furthermore simple per hop pre-ordering on batman-adv's layer shall be implemented to increase the TCP performance. - - -h4. Unit Test Infrastructure - -*Brief description:* Create a flexible unit test infrastructure to increase batman-adv's stability. - -As batman-adv resides in the Linux kernel any crash in batman-adv has severe consequences for the whole system. For instance a deadlock or aceesing invalid memory in batman-adv will freeze the whole system. Typically most embedded routers have a watchdog to restart the device automatically in such situations. However in critical usage scenarios, a reboot of the device might not be tolerable. Therefore a unit test infrastructure shall be created which for one thing tests and verifies single functions and functionalities within the batman-adv code base. And for another it shall stress test the complete software for different virtualized, static and dynamic topologies. - - -h4. IPv6 Stateless Autoconf Gateway Solution - -*Brief description:* Implement a gateway solution for IPv6 stateless autoconfiguration - -batman-adv operates on layer 2, therefore intelligent layer 3 gateway handling is not straightforward. batman-adv currently supports intelligent gateway handling for DHCP (both IPv4 and IPv6), in that it forwards any DHCP discovery or request packet to the best (according to batman's metric) gateway available only, via unicast. However, this is not doable for IPv6 stateless autoconfiguration, as usually not the client is not actively requesting an IP address, but the router is advertising it's gateway capabilities periodically. However, such a stateless configuration is usually more desirable than a stateful configuration as in DHCP in dynamic mesh networks as it allows smoother transitions between subnets and is not as prone to packet loss. Therefore a solution for IPv6 Stateless autoconfiguration shall be implemented. This may be done by exploiting the router preference fields in the router advertisement messages, "RFC4191":https://tools.ietf.org/html/rfc4191 , and/or by limiting the range of router advertisements with the help of batman-adv's broadcast packet's TTL field or by introducing a new TQ limit field. A TQ limit field could be dynamically adapted to minimze the overhead of such router advertisements in case of many gateways but ensuring the availability of the best/closest of all gateways in the mesh network. - - -h4. Improve Broadcast Data Performance - -*Brief description:* Optimize the flooding of broadcast data to reduce overhead and increase its packet delivery ratio. - -So far broadcast data packets are flooded through the whole mesh network. As broadcast packets do not have an ARQ mechanism to reduce packet losses like unicast data transfer has, batman-adv (re)transmits broadcast packets on each hop by default. In dense topologies this can introduce unnecessary overhead, in sparse topologies the packet delivery ratio might be too low. Therefore a smarter mechanism than classic flooding shall be implemented to improve batman-adv's performance for broadcast data packets. This may done taking link qualities in the local neighborhood into account and adjusting the number of rebroadcasts dynamically. Or by introducing "MPR-like":https://en.wikipedia.org/wiki/Multipoint_relay mechanisms. - - -h4. Reduce ARP (/ND) latency and overhead - -*Brief description:* Reduce ARP latency and overhead by implementing an ARP cache. - -batman-adv operates on the link layer, therefore any network protocol (e.g. IPv4/6) needs to perform an MAC address lookup for any IP address through the mesh network first. As ARP packets are broadcast packets, batman-adv simply floods them through the whole mesh network. This can create quite some overhead in large scale mesh networks or high latencies in case of mesh networks with poor links. Therefore an ARP cache shall be implemented on each node so that they could answer ARP requests directly to any host on their segment instead of flooding it through the whole mesh again. Something similiar may be implemented for IPv6's Neighbor Discovery (ND) mechanism. - - -h4. Dead node fast path switching/invalidating - -*Brief description:* Improve batman-adv to respond quicker to node/link failures and avoid packet drops in case of link failures. - -When a node notices the breakdown of a neighbor (see [[routing_scenarios#Convergencespeed|routing scenarios]] to get an idea about the conditions), this node could send any data packet, which it would usually send over this neighbor to either its second best hop if available (which does not always have to be the case due to OGM forwarding policies). Or it could send the packet back to the next hop towards the source again. With the help of sequence numbers, any node on the 'backtracking' path (the backtracking path can be different from the usual path in case of asymmetric links) could notice that a path became invalid very quickly. - - -h4. Link layer FEC/ARQ/Fragmentation module - -*Brief description:* Create a kernel module that can be added onto an interface which performs link layer optimizations (FEC, ARQ, Fragmentation, ...). - -The ARQ and FEC mechanisms provided in 802.11 for unicast data packets is mainly optimized for one hop scenarios only. While the packet delivery ratio on one hop might be fine, it might not be usable for multiple hops in a mesh network. Therefore a link-layer kernel module shall be implemented which performs Forward Error Correction and Automatic Repeat reQuest with dynamic parametrizations, to achieve a certain, configured target packet delivery ratio. Furthermore, the fragmentation from batman-adv could be moved to this link layer module, too. Such a module would then provide a virtual interface which batman-adv would use instead of the actual wifi/ethernet interfaces. - - -h3. Further Tasks - -* Link quality / Packet delivery ratio measurement improvements - -> weighted/exponential moving average - -> testing / performance measurements of / improving NDP, find and improve "performance bottlenecks" -* Multicast Optimization Algorithm enhancements - -> implement reactive tracker packet mode - -> decrease latency of reactive tracker packet mode: attaching small data packets to tracker packet, if MTU fits (otherwise broadcast) in react. mode? ...? - -> ultimately, make proactive tracker mode obsolete - -> optimize number of rebroadcasts (we have the info about the number of next hop nodes) - -> don't send tracker packets if <= 1 destination or > 50% of all nodes are destinations - -> Enhance HNA, host network announcements, improvement scheme to also support MCAs, multicast announcements - -> ... -* Further HNA roaming improvements -* "Network Coding?":https://en.wikipedia.org/wiki/Linear_network_coding -* Packet signing: Add a signing mechanism to identify OGM's sender and drop malicious nodes -* built-in bandwidth test tool ? - - -h2. batctl - -* Enhance live link quality monitoring: bisect -> dot-files -> graphs -> to (live) video? (+ adding horst tool information, wifi interface stats?) -* Add further info to batman-adv vis servers/clients for debugging networks -* bisect, include initial state -* live vis in map (HTML5 + Openstreetmap?) - - -h2. misc - -* multiple interface support in Mesh 3D -* ap51flash GUI; ap51flash multi-flash on single interface -* Android/Maemo/Meego porting + _maintenance_! -* Cooperative work with Pidgin persons? (video/audio/file-transfer/bonjour in pidgin and improving its + batman-adv's combined performance in a mesh network?) \ No newline at end of file diff --git a/open-mesh/Gsoc2011-roadmap.textile b/open-mesh/Gsoc2011-roadmap.textile deleted file mode 100644 index 453e7ce..0000000 --- a/open-mesh/Gsoc2011-roadmap.textile +++ /dev/null @@ -1,15 +0,0 @@ -h1. Google Summer of Code 2011 - Roadmap - -h2. Linus - -* Document NDP-OGM integration -* Document reactive OGM protocol -* Rename NDP -* Adding changes discussed at WBMv4 to NDP -* Rewrite proactive OGM protocol -* Implement reactive OGM protocol (simple) -* Testing -* Reviewing, add feedback -* ... -* (NDP EWMA) -* (reactive OGM protocol, more generic) \ No newline at end of file diff --git a/open-mesh/Gsoc2012-ideas.textile b/open-mesh/Gsoc2012-ideas.textile deleted file mode 100644 index 69bcfc1..0000000 --- a/open-mesh/Gsoc2012-ideas.textile +++ /dev/null @@ -1,9 +0,0 @@ -h1. Gsoc2012-ideas - -We have a page with several ideas/research proposals that you can check ([[ResearchProposals]]). -Feel free to pick one of them and contact us on the mailing-list to discuss about it! - -last years: - -* [[Gsoc2011-ideas]] -* [[Gsoc2010-ideas]] \ No newline at end of file

1 0

[doc] backup-redmine/2020-08-16,master: doc: open-mesh/Kernel_hacking_Debian_image (473e909)
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

Repository : ssh://git@diktynna/doc On branches: backup-redmine/2020-08-16,master >--------------------------------------------------------------- commit 473e909d3b882ef59aa865c1913ffae02392c7fb Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Aug 16 13:29:39 2020 +0000 doc: open-mesh/Kernel_hacking_Debian_image >--------------------------------------------------------------- 473e909d3b882ef59aa865c1913ffae02392c7fb open-mesh/Kernel_hacking_Debian_image.textile | 1 + 1 file changed, 1 insertion(+) diff --git a/open-mesh/Kernel_hacking_Debian_image.textile b/open-mesh/Kernel_hacking_Debian_image.textile index 23ae11e..2eb6bfd 100644 --- a/open-mesh/Kernel_hacking_Debian_image.textile +++ b/open-mesh/Kernel_hacking_Debian_image.textile @@ -285,6 +285,7 @@ CONFIG_PRINTK_TIME=y CONFIG_PRINTK_CALLER=y CONFIG_DEBUG_MISC=y CONFIG_PROVE_RCU_LIST=y +CONFIG_DEBUG_FORCE_FUNCTION_ALIGN_32B=y # for GCC 5+ CONFIG_KASAN=y

1 0

[linux-merge]linux integration; annotated tag, v4.14.193, created. v4.14.193
by postmaster＠open-mesh.org 16 Aug '20

16 Aug '20

The annotated tag, v4.14.193 has been created at 3473746ecd5e0290a9a034fa2abe26d442a60807 (tag) tagging 14b58326976de6ef3998eefec1dd7f8b38b97a75 (commit) replaces v4.14.191 tagged by Greg Kroah-Hartman on Fri Aug 7 09:39:00 2020 +0200 - Shortlog ------------------------------------------------------------ This is the 4.14.193 stable release -----BEGIN PGP SIGNATURE----- iQIzBAABCAAdFiEEZH8oZUiU471FcZm+ONu9yGCSaT4FAl8tBRQACgkQONu9yGCS aT6wuxAAqENTrFxrX4jngMqHiKfUhDWyGzxGaspbBfmb6c3memmY3u1Z0MRJ4XlI G+h5HV6lDeikXLPdp2K1ydCfIZgUA3cxPpCzcAizEtz0YXlVkib8VF9Z3BNgk2WZ vku2otVg2CN/eoBMROvgQkK4WoL39c7ZSJ4MStRC6LIzgUyVc3vnzuUJk5mE3D96 JSIxBaEpXzFf72auHsj2cBV55TaIKLu8xlCe4pcROaqvJtajxTwVBKMg5/XtZWlK qS82xj12O6R/NjPIKKI51/hzO4ibOGDNsuJH4FBR1FDSTwUTSkqKk0Fc+lZZVmZY upytKZJaQNLvmSro9PGrWAUk06xr8GRTuWWegXU+X2/G5ftN0xXuQi4M3xC3g2OQ Z9+asbb/vUnvpYg0e/BgK9v4qt39OwNMIHPeQj9ta89MeMKHl/ObL3EskV+7qJpA tZOxMgfHXHyD+ySH0HSZ5Mx3AaBbVjjhfRsPkRQJE5pYPbWOwFe7snBCnQPcetYT njoXkCDJUC/iabbLyHzRyrEtS97BiWZ4psUDf0RDVsMNFeMF0heBM/Td+eEN6WVo tjZ9HgRKkGWo2+u81RPRclopdsVz/ntWpsKEzjhaTDZFXjfONhs5403yg33IN2Y3 bjixbiH7ErQAT5LFiuE3aSG/WfbhbGsBeEV4qv1di2Vv4Xkpe40= =X7gH -----END PGP SIGNATURE----- ----------------------------------------------------------------------- -- linux integration

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

commits