openmptcprouter
openmptcprouter copied to clipboard
OpenMPTCProuter v0.59 rc3
Changes:
- VPS script Debian 11 support
- MPTCP-next kernel support on app side
- Kernel 5.4 update
- RUTX12 support
- MPTCP redundant fix on 5.4
- SIP ALG is disabled by default on VPS side and option available to enable it in router interface
- Fix multiple server failover
- Speed and duplex settings for each port
- With router using 5.15 kernel, VPS can be 5.4 or 5.15 script
- V2Ray can be used only for TCP (instead of TCP & UDP)
- Update ndpi-netfilter2
- Many fixes (Complete changelog: https://github.com/Ysurac/openmptcprouter-feeds/commits/develop)
I hope that RC3 is the latest Release Candidate...
Previous Beta
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta3-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta3-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta4-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
use
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta4-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta5-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta5-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta6-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta6-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta7-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta7-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta8-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59beta8-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc1-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc1-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc2-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc2-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
5.4 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc3-5.4/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
5.15 kernel:
Images: https://download.openmptcprouter.com/release/v0.59rc3-5.15/
VPS script: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | UPSTREAM="yes" sh
(5.15 kernel based release is tested only on x86_64, only fullmesh (aggregation) is supported, often slower than 5.4, use at your own risk)
May I ask if kernel 5.15 is the one that fixes ipheth packet size problems?
Patch is in both kernel.
Okay, will try 5.4 then since it should be more stable. Thank you for the quick response.
I am getting a user.notice every 4 seconds after I updated to this beta: Mon Feb 21 01:01:29 2022 user.notice post-tracking-post-tracking: Reinstall packages...
Any ideas on what it could be?
VPS script is the same for both?
I used the command listed underneath the 5.4 release: wget -O - http://www.openmptcprouter.com/server-test/debian10-x86_64.sh | sh
the version of VPS says: 0.1026 5.4.100
Okay I ran the command again, and it does say 1027-test now, but I am still getting that spam of "Reinstall packages...". (Yes I did try rebooting my raspberry)
The "Reinstall packages..." is because you did an update and a script try to reinstall packages that was installed. It was not working as I didn't make links on my server for that. I will solve this to check less often.
Hmm, well the messages stopped about 5 hours ago. Were these messages a sign of something bad? Is there anything I need to do?
No, I made the link on the server so packages should be installed. Nothing bad and no real problem here, only too much messages in log.
Ah okay thanks. No iPad wan dying on me so far. Cross your fingers that patch fixed it.
Ignore the post I just made. I just confirmed with pings that a large MTU did not crash my iPad so that means this patch solved one issue. But this now leaves me still confused on why the iPad switches to a permanent down state until I restart it sigh.
Are newer versions going to be without the schedulers such as bbr2?
5.4 release will keep bbr2, for 5.15 I don't know. BBR2 is made by google and for now I don't know current status (git project seems dead).
I run Banana Pi R64 sdcard, after install squashfs go to run and i got:
[ 3.892629] VFS: Cannot open root device "mmcblk1p65" or unknown-block(0,0): error -6 [ 3.900504] Please append a correct "root=" boot option; here are the available partitions: [ 3.908867] b300 62521344 mmcblk0 [ 3.908874] driver: mmcblk [ 3.915662] b301 512 mmcblk0p1 5452574f-2211-4433-5566-778899aabb01 [ 3.915667] [ 3.924624] b302 2048 mmcblk0p2 5452574f-2211-4433-5566-778899aabb02 [ 3.924628] [ 3.933579] b303 1024 mmcblk0p3 5452574f-2211-4433-5566-778899aabb03 [ 3.933583] [ 3.942539] b304 32768 mmcblk0p4 5452574f-2211-4433-5566-778899aabb04 [ 3.942543] [ 3.951499] b305 7168 mmcblk0p5 5452574f-2211-4433-5566-778899aabb05 [ 3.951503] [ 3.960458] b306 524288 mmcblk0p6 5452574f-2211-4433-5566-778899aabb06 [ 3.960462] [ 3.969419] 103:00000 495 mmcblk0p128 5452574f-2211-4433-5566-778899aabb80
mmcblk1p65 - partition not exists - any suggestion ?
·@weblivepro you use 5.4 or 5.15 image ?
I use 5.15
I can't run correct initramfs version from SD card version, something wrong, network not starts, console not work, but i found other sd card bootloader and they running correctly initramfs.
@ImprezyLive for BPI-R64 ? I looked at the code and yes it can't work for now, need some update. 5.15 is not yet tested on anything other than x86_64.
Yes i use BPI-R64.
On this site: https://forum.banana-pi.org/t/bpi-r64-quick-start-boot-from-emmc/9809 i usese SD Card Image And i can run initramfs without any problems, but i can't install any squashfs image.
If i use SD Card Image from you - initramfs not starting correctly, no network, console working in weird mode.
I compile from source (DEVELOP) but without successful.
beta 5 runs great for me, upgrade went smooth both vps and router, x86 router.
feels more solid than 0.57 at least , more consistent speeds when uploading is where i noticed it the most
Fresh installation on vps and rpi4, on the rpi I had to ignore the two preconfigured wans (wan1 and wan2) without deleting them and configuring new ones to use it without problems. If you need more details or logs, tell me what to do. Sorry for my English and thank you for your work.
@evilsep did you create also macvlan or use some USB to ethernet key ? It's the 5.4 image ?
Can you give me the result of uci show network or the part that begin with "network" in System->OpenMPTCProuter, "show all settings" tab ?
@Ysurac I use one usb to ethernet and three smartphones in tethering. All configured without macvlan. The image is 5.4.
The result of uci show network:
network.loopback=interface
network.loopback.proto='static'
network.loopback.ipaddr='127.0.0.1'
network.loopback.netmask='255.0.0.0'
network.loopback.multipath='off'
network.loopback.device='lo'
network.loopback.metric='5'
network.globals=globals
network.globals.multipath='enable'
network.globals.mptcp_scheduler='blest'
network.globals.mptcp_checksum='0'
network.globals.mptcp_debug='0'
network.globals.mptcp_subflows='3'
network.globals.mptcp_add_addr_accepted='1'
network.globals.mptcp_add_addr_timeout='120'
network.globals.mptcp_fullmesh_num_subflows='1'
network.globals.mptcp_fullmesh_create_on_err='1'
network.globals.mptcp_ndiffports_num_subflows='1'
network.globals.ula_prefix='fda7:9093:011b::/48'
network.globals.mptcp_version='0'
network.globals.mptcp_rr_cwnd_limited='Y'
network.globals.mptcp_rr_num_segments='1'
network.globals.mptcp_path_manager='fullmesh'
network.globals.mptcp_syn_retries='8'
network.globals.congestion='cubic'
network.lan=interface
network.lan.proto='static'
network.lan.ipaddr='192.168.100.1'
network.lan.netmask='255.255.255.0'
network.lan.device='eth0'
network.lan.ifname='eth0'
network.lan.ipv6='0'
network.lan.delegate='0'
network.lan.multipath='off'
network.lan.ip4table='lan'
network.lan.metric='6'
network.lan.defaultroute='0'
network.lan.peerdns='0'
network.lan_rule=rule
network.lan_rule.lookup='lan'
network.lan_rule.priority='100'
network.wan1=interface
network.wan1.proto='static'
network.wan1.ip4table='wan'
network.wan1.defaultroute='0'
network.wan1.device='wan1'
network.wan1.masterintf='eth1'
network.wan1.ipv6='0'
network.wan1.metric='3'
network.wan1.peerdns='0'
network.wan1.multipath='off'
network.wan1.addlatency='0'
network.wan3=interface
network.wan3.device='eth1'
network.wan3.ip4table='wan'
network.wan3.defaultroute='0'
network.wan3.metric='7'
network.wan3.peerdns='0'
network.wan3.ipv6='0'
network.wan3.proto='dhcp'
network.wan3.multipath='master'
network.wan4=interface
network.wan4.device='usb0'
network.wan4.ip4table='wan'
network.wan4.multipath='on'
network.wan4.defaultroute='0'
network.wan4.metric='8'
network.wan4.peerdns='0'
network.wan4.ipv6='0'
network.wan4.proto='dhcp'
network.wan5=interface
network.wan5.device='usb1'
network.wan5.ip4table='wan'
network.wan5.multipath='on'
network.wan5.defaultroute='0'
network.wan5.metric='9'
network.wan5.peerdns='0'
network.wan5.ipv6='0'
network.wan5.proto='dhcp'
network.wan1_dev=device
network.wan1_dev.name='wan1'
network.wan1_dev.type='macvlan'
network.wan1_dev.ifname='eth1'
network.wan1_dev.mode='vepa'
network.wan1_dev.macaddr='a0:95:d8:6e:a1:d6'
network.wan2=interface
network.wan2.proto='static'
network.wan2.ip4table='wan'
network.wan2.defaultroute='0'
network.wan2.device='wan2'
network.wan2.masterintf='eth1'
network.wan2.ipv6='0'
network.wan2.metric='4'
network.wan2.peerdns='0'
network.wan2.multipath='off'
network.wan2.addlatency='0'
network.wan2_dev=device
network.wan2_dev.name='wan2'
network.wan2_dev.type='macvlan'
network.wan2_dev.ifname='eth1'
network.wan2_dev.mode='vepa'
network.wan2_dev.macaddr='40:76:93:65:e8:52'
network.omrvpn=interface
network.omrvpn.device='tun0'
network.omrvpn.ip4table='vpn'
network.omrvpn.multipath='off'
network.omrvpn.leasetime='12h'
network.omrvpn.type='tunnel'
network.omrvpn.txqueuelen='100'
network.omrvpn.metric='1200'
network.omrvpn.proto='none'
network.omr6in4=interface
network.omr6in4.proto='6in4'
network.omr6in4.ip4table='vpn'
network.omr6in4.multipath='off'
network.omr6in4.ipaddr='10.255.255.2'
network.omr6in4.peeraddr='10.255.255.1'
network.omr6in4.auto='0'
network.omr6in4.metric='1201'
network.omr6in4.ip6addr='fe80::a00:2/126'
network.omr6in4.gateway='fe80::a00:1/126'
network.lan_dev=device
network.lan_dev.name='eth0'
network.wan3_dev=device
network.wan3_dev.name='eth1'
network.wan4_dev=device
network.wan4_dev.name='usb0'
network.wan5_dev=device
network.wan5_dev.name='usb1'
network.wan6=interface
network.wan6.proto='dhcp'
network.wan6.device='usb2'
network.wan6.multipath='on'
network.wan6.addlatency='0'
network.wan6.defaultroute='0'
network.wan6.peerdns='0'
network.wan6.ipv6='0'
network.wan6.metric='11'
Is this the VPS version I should be using for beta 5? Version 0.1027-test 5.4.132-mptcp
Also in the past few betas I am seeing that my DSL gets bad ping on the status page. Before I push too much against Vultr (my VPS provider) I would like to know if there is any reason the VPS script would be causing extra random latency?
In the meantime I have tried making my own OMR-tracker that pings only the gateway, but it refuses to ping my ISP and continues to send pings to the server.
edit when I do ifconfig I see a list of ovpnbonding going from 1-8. Is this normal?
edit 2 Looks like it was a random hop between the server and my ISP that had gone bad. You can delete this comment unless you think answering my other questions would be educational for others.
Just reporting that I can't put port 80 into OMR bypass without it counting down to 0 and saying "do you want to revert?"
I thought I could be cheeky and instead add http as a service but while it saved with no errors, it eventually caused luci to lock up and I could no longer access the page (ssh was fine though). Curious why it doesn't like port 80 being bypassed?
@darthclide you mean in omr-bypass source port ? This conflict with OMR if you use port 80. You can still use port 443 to connect to web interface.
Hmmm, what if I need to redirect both port 80 and 443 to a specific interface? Am I out of luck?
You can change OMR port if you want.
Ah well I managed to find a way around my issue. So I think it best for me not to change something you have set by default.