commits July 2017

commits@lists.open-mesh.org

1 participants
1317 discussions

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Bridge-loop-avoidance-II (35d10d88)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Bridge-loop-avoidance-II (adad2d42)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

1 0

[doc] backup-redmine/2017-07-13, master: doc: open-mesh/Filing (fe22a416)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Understand-your-batman-adv-network (465517f8)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit 465517f89295bbfad600c35b78ac246d18f27ff7 Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Mon Aug 22 01:34:15 2011 +0000 doc: batman-adv/Understand-your-batman-adv-network >--------------------------------------------------------------- 465517f89295bbfad600c35b78ac246d18f27ff7 batman-adv/Understand-your-batman-adv-network.textile | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/batman-adv/Understand-your-batman-adv-network.textile b/batman-adv/Understand-your-batman-adv-network.textile index b76d5ef1..b6f62c3c 100644 --- a/batman-adv/Understand-your-batman-adv-network.textile +++ b/batman-adv/Understand-your-batman-adv-network.textile @@ -37,11 +37,13 @@ The local translation table (mac addresses announced by this host) can be found <pre> cat /sys/kernel/debug/batman_adv/bat0/transtable_local -Locally retrieved addresses (from bat0) announced via HNA: -* fe:fe:00:00:01:01 +Locally retrieved addresses (from bat0) announced via TT (TTVN: 1): + * fe:fe:00:00:01:01 </pre> -Every batman node announces at least one mac address - the mac of the batX interface. +The current translation table state is represented by the tt version number that is propagated in the mesh. + +Note: Every batman node announces at least one mac address - the mac of the batX interface. The global translation table (mac addresses announced by other hosts) can be found in the transtable_global file:

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Tweaking (35678b08)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit 35678b08e741c47ba3cae155b8ac68a25d0175ff Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Mon Aug 22 01:06:17 2011 +0000 doc: batman-adv/Tweaking >--------------------------------------------------------------- 35678b08e741c47ba3cae155b8ac68a25d0175ff batman-adv/Tweaking.textile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/batman-adv/Tweaking.textile b/batman-adv/Tweaking.textile index 697fb810..88be2246 100644 --- a/batman-adv/Tweaking.textile +++ b/batman-adv/Tweaking.textile @@ -142,7 +142,7 @@ h3. log level Available since: batman-adv 2010.1.0 (if debugging has been compiled-in) -The standard warning and error messages which help to setup & operate batman-adv are sent to the kernel log. However, batman-adv also offers extended logs that can be used to understand and/or debug the routing protocol. Keep in mind that it is necessary to activate debugging at compile time before you can use these facilities (consult "the README file":http://www.open-mesh.org/projects/batman-adv/repository/revisions/master/entry/README to learn how to set the compile option). Per default, the logging is deactivated (log level: 0), Enable log level '1' to log messages related to routing / flooding / broadcasting. Log level '2' only shows messages related to route or hna added / changed / deleted. Switch to log level '3' to get all log messages. +The standard warning and error messages which help to setup & operate batman-adv are sent to the kernel log. However, batman-adv also offers extended logs that can be used to understand and/or debug the routing protocol. Keep in mind that it is necessary to activate debugging at compile time before you can use these facilities (consult "the README file":http://www.open-mesh.org/projects/batman-adv/repository/revisions/master/entry/README to learn how to set the compile option). Per default, the logging is deactivated (log level: 0), Enable log level '1' to log messages related to routing / flooding / broadcasting. Log level '2' only shows messages related to route added / changed / deleted. Log level '4' focuses on translation table operations. Switch to log level '7' to get all log messages. <pre> cat /sys/class/net/bat0/mesh/log_level

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Understand-your-batman-adv-network (8e5e612d)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit 8e5e612d6afa56474cf49aca09077228459288bb Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Mon Aug 22 01:40:20 2011 +0000 doc: batman-adv/Understand-your-batman-adv-network >--------------------------------------------------------------- 8e5e612d6afa56474cf49aca09077228459288bb batman-adv/Understand-your-batman-adv-network.textile | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/batman-adv/Understand-your-batman-adv-network.textile b/batman-adv/Understand-your-batman-adv-network.textile index b6f62c3c..a6be93d0 100644 --- a/batman-adv/Understand-your-batman-adv-network.textile +++ b/batman-adv/Understand-your-batman-adv-network.textile @@ -42,7 +42,6 @@ Locally retrieved addresses (from bat0) announced via TT (TTVN: 1): </pre> The current translation table state is represented by the tt version number that is propagated in the mesh. - Note: Every batman node announces at least one mac address - the mac of the batX interface. @@ -50,11 +49,12 @@ The global translation table (mac addresses announced by other hosts) can be fou <pre> cat /sys/kernel/debug/batman_adv/bat0/transtable_global -Globally announced HNAs received via the mesh bat0: -* fe:fe:00:00:01:01 via fe:fe:00:00:02:01 +Globally announced TT entries received via the mesh bat0 + Client (TTVN) Originator (Curr TTVN) + * fe:fe:00:00:01:01 ( 1) via fe:fe:00:00:02:01 ( 1) </pre> -Each line contains the announced mac address and the mac of the batman node announcing it. +Each line contains the announced mac address plus the ttvn at the point of the addition to the table and the mac of the batman node announcing it including the latest ttvn of that very node. h3. Gateway table

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Release-todo (3fff669f)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit 3fff669fd50c8aaffc516ca0212066c88d32bedb Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Mon Aug 22 11:55:01 2011 +0000 doc: batman-adv/Release-todo >--------------------------------------------------------------- 3fff669fd50c8aaffc516ca0212066c88d32bedb batman-adv/Release-todo.textile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/batman-adv/Release-todo.textile b/batman-adv/Release-todo.textile index d336c21e..58559c1e 100644 --- a/batman-adv/Release-todo.textile +++ b/batman-adv/Release-todo.textile @@ -28,7 +28,7 @@ h2. Documentation * [x] build changelog * [x] update README * [x] write release announcement -* [ ] update wiki documentation +* [x] update wiki documentation h2. Release @@ -42,7 +42,7 @@ h2. Release h2. Distributions -* [ ] update openwrt package +* [x] update openwrt package * [ ] update debian package

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/OGMv2 (bf48f391)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit bf48f391c3c0f36bf9e650964b4c5476b6b2024e Author: Simon Wunderlich <sw(a)simonwunderlich.de> Date: Wed Sep 28 18:33:25 2011 +0000 doc: batman-adv/OGMv2 >--------------------------------------------------------------- bf48f391c3c0f36bf9e650964b4c5476b6b2024e batman-adv/OGMv2.textile | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-) diff --git a/batman-adv/OGMv2.textile b/batman-adv/OGMv2.textile index 0eed8d9f..649a5ffd 100644 --- a/batman-adv/OGMv2.textile +++ b/batman-adv/OGMv2.textile @@ -120,21 +120,21 @@ The following steps need to be undertaken in the updating and creating process: h4. 5.2. Purging Outdated Router Entries -It might happen that for instance from a certain Neighbor we would receive an OGM of perfect quality first and will chose that Router. However after that the path over that Selected Router could suddenly be jammed, leading to no more updates from that Router, resulting in a stale entry. Newer and newer (higher sequence number) OGMs might arrive over other potential Routers, but would never be chosen because of a path TQ never being better than perfect, highest path TQ of the currently Selected Router. Therefore we need to at some point consider this Selected Router as outdated and switch to one of the alternative, loop-free Routers in our list which provide more up-to-date information. This is not being done time-based but based on the sequence number, a Selected Router may only be chosen if its OGM has not been older than OGM_SEQ_RANGE sequence numbers. +It might happen that for instance from a certain Neighbor we would receive an OGM of perfect quality first and will chose that Router. However after that the path over that Selected Router could suddenly be jammed, leading to no more updates from that Router, resulting in a stale entry. Newer and newer (higher sequence number) OGMs might arrive over other potential Routers, but would never be chosen because of a path TQ never being better than perfect, highest path TQ of the currently Selected Router. Therefore we need to at some point consider this Selected Router as outdated and switch to one of the alternative Routers in our list which provide more up-to-date information. This is not being done time-based but based on the sequence number, a Selected Router may only be chosen if its OGM has not been older than OGM_SEQ_RANGE sequence numbers. -Note that a lower OGM_SEQ_RANGE favours chosing Routers with the most-up-to date information: This especially penalizes asymmetric links and paths - although we do not receive that many OGMs from such a Router with such an asymmetric path (showing a not that good receive quality), it might still be the best choice for transmitting our own data packets though. And could lead to fast route flapping also in symmetric topologies when OGMs in general have a low probability of arrival. However having a too large OGM_SEQ_RANGE might favour too old, outdated information too much, as described with the example before. +Note that a lower OGM_SEQ_RANGE favors choosing Routers with the most-up-to date information: This especially penalizes asymmetric links and paths - although we do not receive that many OGMs from such a Router with such an asymmetric path (showing a not that good receive quality), it might still be the best choice for transmitting our own data packets though. And could lead to fast route flapping also in symmetric topologies when OGMs in general have a low probability of arrival. However having a too large OGM_SEQ_RANGE might favor too old, outdated information too much, as described with the example before. More precisely we have to: * If the OGM's Sequence Number is newer than the Originator's Sequence Number: ** The new Originator's Sequence Number must be set to the Sequence Number contained in the received OGM. -** for all Routers of the OGM's originator: if (Originator's Sequence Number - Router's Sequence Number) > OGM_SEQ_RANGE, purge the router from the OGM's originator's Router List. +** for all Routers of the OGM's originator: if the Originators Sequence Number is by OGM_SEQ_RANGE newer than the Routers Sequence Number, purge the router from the OGM's originator's Router List. Note that this step can result in rebroadcasting an OGM in step 5.3. which is not the one we have actually received and are currently processing - which is intended: This incoming OGM might be the cause of purging outdated entries, however there might be still other Routers in the Router list which have a higher path TQ and are therefore more desirable to chose as the new Selected Router than the Router we received this OGM from. h4. 5.3. Switching to (or Keeping) best Router -This step ensures a good Router selection to the best knowledge of a node. As the Router list only keeps potential, loop-free nodes (due to steps 4.2.2 and 5.4) which are further not too old (due to step 5.2) we can now freely choose any node from this list. If in this round we got an OGM of a Router which we did not and will not chose as the Selected Router (due to a lower path TQ, although it would be feasible to chose it due to a newer sequence number of this OGM), than we just updated this Routers values, without selecting it. Instead the next steps will chose the same old Router (which is not the one we received the current OGM from) again - but will avoid rebroadcasting the OGM of the old, though still best old and newly Selected Router, due to the Rebroadcasted flag. +This step ensures a good Router selection to the best knowledge of a node. The Routers values are updated regardless whether it is the selected router or not. If the Rebroadcasted Flag is set for the selected router and the OGM was not received by this router, the OGM will not be rebroadcasted. Specifically, we must undertake the following actions: @@ -145,31 +145,22 @@ Specifically, we must undertake the following actions: h4. 5.4. Purging non-potential Routers -When rebroadcasting a new OGM certain other Routers do not guarantee loop-freeness anymore. We can still chose the Routers from our list that either have broadcasted a higher sequence number than the one we might have just rebroadcasted, they or any next hop behind them are not allowed to switch their route to a lower sequence number (like the one we might just have rebroadcasted) again. Or we could chose any router in the list with the same sequence number and a higher path TQ than the one of the Selected Router (though this will not be the case, because step 5.3. ensured that we are already chosing the Router with the highest path TQ from our Router list). In all other cases we do not consider these Neighbors as potential Routers anymore, they or any next hop behind them might start chosing us as their router without us noticing. Therefore we remove them from the list to ensure a safe Router list: +To keep the routers list short and clean, we will purge routers with old sequence numbers which are not the selected router or have a lower TQ. These Routers are obsolete and will not be chosen in the future. If an OGM was rebroadcasted in the previous step: * Purge all routers with a sequence number smaller than the Selected Router's Sequence Number. * Purge all routers with a sequence number equal to the Selected Router's Sequence Number and a path TQ smaller than the Selected Router's path TQ. -Note: If no OGM was rebroadcasted in the previous section then no purging will be done in this section anyways. However the "If" shall emphasize, that not the switching of the Selected Router makes the router list clean-up in this section mandatory to ensure loop-freeness, but the rebroadcasting of an OGM does. +Note: If no OGM was rebroadcasted in the previous section then no purging will be done in this section anyways. (TODO: can we skip this condition?) h3. 6. Re-broadcasting other nodes' OGMs - When an OGM is to be re-broadcasted some of the message fields must be changed others must be left unchanged. All fields not mentioned in the following section remain untouched: * The TTL must be decremented by one. If the TTL becomes zero (after the decrementation) the packet must be dropped. -* The hop penalty must be applied on the OGM's TQ field. See 'Penalties - Hop Penalty' for further details. If the OGM's TQ becomes zero (after hop penalty) the packet must be dropped. - -h2. 7. Penalties - -h3. 7.1 Hop Penalty - -In certain network setups the link quality between neighbors is very similar whereas the number of hops is not. In these scenarios it is desirable to chose the shortest path to reduce latency and to safe bandwidth (especially on wireless mediums). The hop penalty is a value greater than zero and smaller or equal to one. It is a fixed value but may be changed during runtime. The hop penalty is applied on an outgoing OGM in the followig way: - -* Outgoing OGM's TQ = path TQ * (1 - hop penalty) +* The hop penalty must be applied on the OGM's TQ field. If the OGM's TQ becomes zero (after hop penalty) the packet must be dropped. The hop penalty is applied by reducing the path TQ by HOP_PENALTY percent. -The result always needs to be rounded down to ensure that an outgoing OGM's TQ is always smaller than the incoming OGM's TQ. +Note: The Hop penalty is especially required in loss-less networks where shorter path should be favored. It is also useful to select the shortest path in lossy networks where the TQ values are very similar. h2. 8. Proposed Values for Constants

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/SubmittingLinux (173fd328)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit 173fd3289db112b8b484a5842cf71e73ba9d2223 Author: Marek Lindner <mareklindner(a)neomailbox.ch> Date: Thu Sep 8 16:27:01 2011 +0000 doc: batman-adv/SubmittingLinux >--------------------------------------------------------------- 173fd3289db112b8b484a5842cf71e73ba9d2223 batman-adv/SubmittingLinux.textile | 2 ++ 1 file changed, 2 insertions(+) diff --git a/batman-adv/SubmittingLinux.textile b/batman-adv/SubmittingLinux.textile index 0f4dd299..e12c1e35 100644 --- a/batman-adv/SubmittingLinux.textile +++ b/batman-adv/SubmittingLinux.textile @@ -91,6 +91,8 @@ rm 'net/batman-adv/compat.h' [linux f2204f0] Merge remote branch 'origin/standalone/next' into merge/master </pre> +All changes to the the sysfs-class-net-* files can't be automatically merged because they reside in a different subfolder. As a consequence the new version has to be manually copied from a local git tree to 'Documentation/ABI/testing/'. The README file also requires manually handling: All relevant in-kernel changes have to be merged into 'Documentation/networking/batman-adv.txt' whereas out-of-kernel information has be stripped out. + h2. check for compat code

1 0

[doc] backup-redmine/2017-07-13, master: doc: batman-adv/Release-todo (cd9e6c8e)
by postmaster＠open-mesh.org 14 Jul '17

14 Jul '17

Repository : ssh://git@open-mesh.org/doc On branches: backup-redmine/2017-07-13,master >--------------------------------------------------------------- commit cd9e6c8e54ebd7ac0acfb77d406a91b620e86ec7 Author: Sven Eckelmann <sven(a)narfation.org> Date: Mon Oct 3 02:54:48 2011 +0000 doc: batman-adv/Release-todo >--------------------------------------------------------------- cd9e6c8e54ebd7ac0acfb77d406a91b620e86ec7 batman-adv/Release-todo.textile | 42 ++++++++++++++++++++--------------------- 1 file changed, 21 insertions(+), 21 deletions(-) diff --git a/batman-adv/Release-todo.textile b/batman-adv/Release-todo.textile index 58559c1e..8bd418f2 100644 --- a/batman-adv/Release-todo.textile +++ b/batman-adv/Release-todo.textile @@ -14,41 +14,41 @@ h2. Release Scheduling h2. Code -* [x] make sure all code contains the release version strings -* [x] update copyright years (if needed) -* [x] remove dead code (if any) -* [x] check against kernel versions -* [x] check that maint branch is merged in next -* [x] check for open patches on the ml -* [x] git tag release +* [ ] make sure all code contains the release version strings +* [ ] update copyright years (if needed) +* [ ] remove dead code (if any) +* [ ] check against kernel versions +* [ ] check that maint branch is merged in next +* [ ] check for open patches on the ml +* [ ] git tag release h2. Documentation -* [x] build changelog -* [x] update README -* [x] write release announcement -* [x] update wiki documentation +* [ ] build changelog +* [ ] update README +* [ ] write release announcement +* [ ] update wiki documentation h2. Release -* [x] build package tar.gz and generate checksums -* [x] move packages/checksums to stable/sources / create symlinks from releases -* [x] update download links in the redmine downloads page -* [x] send release mail -* [x] update IRC channel topic - +* [ ] build package tar.gz and generate checksums +* [ ] move packages/checksums to stable/sources / create symlinks from releases +* [ ] update download links in the redmine downloads page +* [ ] send release mail +* [ ] update IRC channel topic h2. Distributions -* [x] update openwrt package +* [ ] update openwrt package * [ ] update debian package h2. Aftershow -* [x] Merge next into master branch (to get changes on README, CHANGELOG, ...) -* [x] Merge current git tag into maint (must be fast-forward or pre-tagging steps were done wrong) -* [x] Merge patches from master into next (to get all the fancy new features) +* [ ] Merge next into master branch (to get changes on README, CHANGELOG, ...) +* [ ] Merge current git tag into maint (must be fast-forward or pre-tagging steps were done wrong) +* [ ] Merge patches from master into next (to get all the fancy new features) * [ ] Rebase patches in next on top of linux-next and send it to kernel maintainers for future kernels ([[SubmittingLinux]]) +* [ ] re-enable next check in checkstuff.sh

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

commits July 2017