Already merged into master branch:
https://github.com/accel-ppp/accel-ppp/pull/115
- Queries
- All Stories
- Search
- Advanced Search
- Transactions
- Transaction Logs
Advanced Search
Mar 15 2024
Mar 4 2024
Dec 6 2023
Dec 1 2023
Nov 30 2023
Nov 7 2023
Okey, so in that case, how do you deal with it?
this is not related to accel, because accel-ppp is only control-plane for pppoe sessions so slow performance of pppoe is related to kernel, not to accel-ppp
Only difference is type of accel session. ipoe vs pppoe. So this is not related to accel?
This question is not related to accel, traffic forwarding is done by kernel, not by accel-ppp so please contact netdev (https://lore.kernel.org/netdev/ ) if you want to discuss it
Nov 3 2023
Hi @Dimka88, I have implemented this feature for personal needs, but I would like to contribute with the code if possible. Please let me know how can I upload the changes so you can review. Should I do this on Github?
Oct 3 2023
Summarize:
Issue appears with veth and vlan_mon only on 4.19. If use bridge instead, all will work properly. As Deb10 support expired around 1 year we make decision to delete tests on Deb10 from github test
I can confirm that kernel 5.10.0-0.deb10.24-amd64 is fixing issue on Debian 10.
Apr 16 2023
Jan 30 2023
Dec 21 2022
Dec 20 2022
It is not fixed in actual version. It is required to fix here: https://github.com/accel-ppp/accel-ppp/blob/2b865db72bc2ddc6411950d72f1c23e8ef115b8a/accel-pppd/ppp/ipcp_opt_ipaddr.c#L100 , https://github.com/accel-ppp/accel-ppp/blob/2b865db72bc2ddc6411950d72f1c23e8ef115b8a/accel-pppd/ppp/ipv6cp_opt_intfid.c#L209 , https://github.com/accel-ppp/accel-ppp/blob/2b865db72bc2ddc6411950d72f1c23e8ef115b8a/accel-pppd/ctrl/ipoe/ipoe.c#L668
Dec 8 2022
Nov 18 2022
Attached the patch i made for the version 1.12.0-202-g28fe4de. It will use Calling Number/Called Number as RADIUS Calling-Station-Id/Called-Station-Id in case they are present in L2TP ICRQ, otherwise old beahvior is used (LAC/LNS IP addresses).
Nov 15 2022
Oct 5 2022
I think it is not implemented yet, but is is good feature request
Oct 3 2022
Sep 16 2022
just to discuss the possibilities.
I see Calling-Number/Called-Numer AVPs can be taken from attr->val.octets doing something within l2tp_recv_ICRQ procedure:
Sep 5 2022
Aug 15 2022
This cmake gives working deb on VyOS on Debian11 machine:
cmake -DCMAKE_BUILD_TYPE=Debug \
-DCMAKE_C_FLAGS='-g -O0' \ -DBUILD_IPOE_DRIVER=TRUE \ -DBUILD_VLAN_MON_DRIVER=TRUE \ -DKDIR=/usr/src/linux-headers-`uname -r` \ -DCMAKE_INSTALL_PREFIX=/usr \ -DLUA=TRUE \ -DLUA=5.3 \ -DCPACK_TYPE=Debian11 ..
make
cpack -G DEB
i have testsed last version accel. its ok
Hi.
Is there any recommendations on compiling and embedding newly compiled version of accel-ppp to vyos 1.4 ?
I mean how to install newly built package into VyOS 1.4.
The deb built on Debian11 and installed with dpkg on VyOS does not start for some reason with no output.
Aug 4 2022
В логе должны отображаться команды полученые при accel-cmd *.
Необходимо указывать
[cli] versobe=2
Aug 2 2022
patch under test
diff --git a/accel-pppd/ctrl/sstp/sstp.c b/accel-pppd/ctrl/sstp/sstp.c index f17db43..40c6ad9 100644 --- a/accel-pppd/ctrl/sstp/sstp.c +++ b/accel-pppd/ctrl/sstp/sstp.c @@ -2415,12 +2415,12 @@ static int sstp_connect(struct triton_md_handler_t *h)
Jul 28 2022
Is there any chance to fix that ?
Jul 26 2022
build command
wget https://github.com/accel-ppp/accel-ppp/archive/refs/heads/master.tar.gz -O accel-ppp-master.tar.gz
tar -xvf accel-ppp-master.tar.gz
cd accel-ppp-master
mkdir build
cd build
cmake -DCMAKE_INSTALL_PREFIX=/usr -DKDIR=/usr/src/linux-headers-uname -r -DRADIUS=TRUE -DSHAPER=TRUE -DLOG_PGSQL=FALSE -DLUA=TRUE -DBUILD_IPOE_DRIVER=FALSE -DCPACK_TYPE=Debian10 -DBUILD_VLAN_MON_DRIVER=FALSE -DCMAKE_BUILD_TYPE=Debug -DCMAKE_C_FLAGS='-g -O0' ..
make -j $CPU_COUNT.
cpack -G DEB
Add coredump file
Jun 20 2022
added
May 20 2022
May 13 2022
Apr 27 2022
Apr 21 2022
Hi
Any one for help to make this ?
Mar 31 2022
Mar 10 2022
Hi Dimka88!
Feb 17 2022
Jan 30 2022
Jan 26 2022
Jan 9 2022
Jan 5 2022
also, we have to check driver performance when ipoe_lookup_hwaddr is used for many sessions (many sessions in vrf)
It will not work if there in the conditions below:
- two sessions in L2 ipoe shared mode
- the same interface (e.g. eth0)
- first session without vrf
- second session with vrf
- the same ip.src for both sessions
additional information and patches how it should works
Jan 4 2022
But for Service name i need your help to extract from PADI or PADO
Jan 3 2022
i will try to make patch for this.
many of use case send to radius clear mac for user and mac for machine and not need to send interface
this with interface and service will use only fkr cli interface to show local in system .
I use this one: show sessions match calling-sid vlan123
calling-sid/called-sid send to radius as info for user
separate column need only for statistic local in cli when statistic or only to see user per interface
why do you want to have "uplink interface" as a separate column? why calling-sid/called-sid with interface name is not enough?
Hi @svlobanov
Yes this list interface here :
@micron
regarding to interface:
ifname-in-sid=called-sid|calling-sid|both:
Hi
yes check in man/doc but calling-sid or called-sid not get interface .
You can include "uplinkif" to calling-sid or called-sid, it is already implemented. Please check man/doc related to this feature
Dec 24 2021
Related task https://phabricator.accel-ppp.org/T56
Dec 22 2021
Dec 21 2021
Thank you @svlobanov .
Added possibility to define netlink buffers
[common] nl-snd-buffer=32768 nl-rcv-buffer=1048576
Dec 20 2021
Dec 17 2021
Unfortunately, I can not reproduce this issue, please reopen this ticket when this issue will appear again.
no in this case work with latest master and after this segfault not see this problem.
may be is network problem send wrong package
Hi @micron , did you modify the accel-ppp code by yourself, exactly functions related to Service-Name?
Hi @slima , do you have a chance to test on the fresh accel-ppp version from the master branch?
please check this patch https://github.com/accel-ppp/accel-ppp/pull/31
Should be already fixed by https://github.com/accel-ppp/accel-ppp/commit/ff91c7376f2a6200be7ea90d36411e1acb4cd9c8
Let me know if you will face a similar backtrace with the fresh accel-ppp version.
@djan I believe there is no interesting stat from the connlimit module. Maybe you think that it works in a bit different way