I‘d rather suggest building a test iso as we build the frr source code from their repo directly. I would not permanently add this not yet mainlined patchset to the package feed
- Queries
- All Stories
- Search
- Advanced Search
- Transactions
- Transaction Logs
All Stories
Jul 31 2019
Can we make a nightly with the patches from:
Jul 30 2019
Details of the mechanism, introduced after jessie, are in
https://github.com/vyos/vyatta-op-quagga/pull/3 I hope now it looks better
Did I something wrong with https://phabricator.vyos.net/T1550 and https://github.com/vyos/vyatta-op-quagga/pull/2 that the pull request is not showing in phabricator?
Jul 29 2019
I've fixed the setup and re-populated the equuleus repo.
https://github.com/FRRouting/frr/pull/4742 let's give it a try?
Jul 28 2019
Jul 27 2019
Confirm fixed in latest rolling 201907270337
Confirm fixed in latest rolling 201907270337
Jul 26 2019
prototype successfully tested
That error will disappear once https://github.com/vyos/ppp-upstream/pull/1 has been merged in. You can meanwhile try to remove the route option and see if it makes any difference.
All of VyOS is VMs hosted on ESXI
Yeah, I'd mark this not-a-bug. Update the documentation to mention hosts file update with failover doesn't maintain consistent state between failover servers.
Sorry about the delays, been a couple hectic days at work.
Jul 25 2019
@mustard Any updates?
Attached are the pcap and debug logs from a simple setup as outlined above, two hosts. "Master" distributing the route.
I could reproduce it. The DHCP client script overwrites those entries.
Some Feedback from the frr people:
Thanks - but still there should not be a Python error and also a user error message
Got it to work correctly by adding the address of the listen-interface onto eth0
Jul 24 2019
currently testing internally, will be released soon to the current branch.
With the new versions of the open-vm-tools you can explicitly disable the pulling of the routing-table. Maybe that's a better way then disabling the whole pulling of information.
Tested rolling from yesterday and 1.2.2, no issues.
@c-po Still not fully resolved on Crux . Maybe candidate for Backlog on VyOS 1.2.3? Thanks.
@mustard Your ppp session looks just fine. The removal of your default route doesn't terminate the pppoe connection, at least there was no termination request in the dump. Your icmp echo request time out since you don't have a default gateway, so we need to find out why the default gateway gets removed. When you see the icmp timeouts for 8.8.8.8 your pppoe session is still alive and is answering the LCP requests successful, so I think we need to focus on frr and why it is removing the efault route.
The following log lines are produced when the disconnect occurs
Hello @mustard . Can you provide logs when route deleted?
show log tail 50
I enabled debug in /etc/ppp/options and all seems to be ok with the underlying PPPoE connection but I still loose the connection. I think the default route is being dropped some how?
Thanks for trying to reproduce the issue,
Jul 23 2019
personally I like the below most. it will require quite an amount of work for migration.
See T1331
backported
has been fixed by reverting to 4.1 since pdns stopped supporting debian jessie.
Run into the same with 1.2.2