Hello,
I’m running the rc3 on 2011UiAS-2HnD and when I use the webfig to configure BGP templates or connections, I always get error „Invalid value in AS“ no matter what I fill in the AS field. Tested on several web browsers (Firefox 78.10.0esr (64-bit), Chromium Version 90.0.4430.93 (Developer Build) built on Debian 10.9, running on Debian 10.9 (64-bit))
I do not use WinBox, and when I try to configure it from the CLI, everything goes fine.
Interesting, I’ve unpacked another cAP and this time it has 128MB RAM, also has capacitors from diffrent manufacturer.
That cAP with 256MB i’ve bought shortly after release so it looks like true.
Later I will unscrew board from enclosure and see if its IPQ-4018 or IPQ 4019-based.
It is still release candidate, so don’t use it on important or production device.
So many times…
Do not use backup, use export for configuration. Of course, if you have certificates in ex. for OVPN you have to export and store those files too( I mean on computer, not same device because if something go wrong netinstall will format flash).
I’ve upgraded sereval test devices from 6.48.4 to 7.1rc2 and rc3 and all configuration was been converted by crossfig
Thing is, simple export/import is broken in 7.1rc3,
“expected end of line error” when line-break inside double-quote, example :
/user group
set full policy="local,telnet,ssh,ftp,reboot,read,write,policy,test,winbox,pas\
sword,web,sniff,sensitive,api,romon,dude,tikapp,rest-api"
Moreover, bridge filter rules -seem- randomly broken again from rc3.
Didn’t test extensively as connectivity has to be restored.
Will try again with rc4.
Lost access to a CHR after reboot. Turned out that the CHR had lost all config including users so I needed to login with admin and blank password.
So losing config is still an issue. I will see if I can reproduce the problem again but this happens after it had been running for a few days.
Yeah this is major bug in rc3 half of my scripts stoped working due “expected end of line error” had to remove spaces where ever i could and remove “” as temp fix..
Automated tests on real hardware would be great on their side. Like a pool of routers that automatically gets wiped, reconfigured and tested automatically remotely. That would catch so many regressions.
mAP2nD 7rc3
Took me a couple of times to realize what was happening but it seems sometimes it looses PART of the config after a power off/on cycle.
This time it was DNS settings and default route to 0.0.0.0/0 (rather trivial…but also very easy to detect).
Restored from backup from 2 days ago and ok again.
But this happened now 3 times since saturday on this device. Only now I see what’s happening.
23:16:52 script,info WG Peer toggled
23:17:00 system,info led trigger changed
Which brings me to something else I noticed.
I need to disable the peer and enable it again before WG interface kicks in action.
EDIT:
Give it some time and it comes through.
What is the current progress on MPLS L3 / BGP / VRF routing on v7.1?
In combination with native Hardware Offloading and IPv4 / IPv6 support this would be a bummer.
Maybe you are better off starting two different development threads.
One renewing all major existing network connectivity features using new kernel architecture and optimized code (example: /routing/bgp)
Another focusing on new features, like implementing Docker container support
Can someone help me fix my OSPF configuration below that works fine for v6 but doesnt work right in v7? I have a feeling it has something to do with the new filtering rules, but I cannot get it to work properly (so I downgraded to the v6 stable at the moment).