Hi all. Long time lurker/learner, first time I decide to register and ask for help. Hope someone can help me.
I decided to get rid of my IPv6 or Dual Stack setup so I deleted everything from cisco RV-325 router, Proxmox PVE host and all LxC and VM (PBS is a virtual machine BTW. All of them are ok but except 2 guests and now I see the same behaveour on the PVE host.
Issue is: I delete the IPv6 addr with ip addr del 2001:blabla/64 dev eth0 or ens0p18 and it gets deleted but like 5 or 10 seconds later ip addr show shows it again. it happens on 1 LxC that uses vmbr1 and 1 VM that uses vmbr2 , the resto of containers and VM it does not happen.
By I deleted ipv6 everywhere I mean through the PVE GUI all the time and I have confirmed that on all /etc/network/interfaces there is no mention to IPv6. also the cisco RV-325 is no longer dual stack and I even deleted my HE.com IPv6 tunnel just in case.
On the host PVE itself, vmbr1 and vbmr2 are manual and empty like no IP (I use vbmr0 and 3 as the host own dedicated interface and those are static ) so all that seems ok I think. But like I said after deleting that IPv6 on even the vbmr1 and 2 they come back as well right away but it says tentatively first and then dynamic. Even named on syslog mentions that is listening on that Ipv6 yet on its bind9 config file I have commented with // the line to enable listening on v6 .
On my cisco router I have also double check is no longer dual stack, no longer has a tunnel, and it is "supposedly" the only DHCP in my network or setup. I even shutdown all containers and VM except PBS and the PVE host it self that has it as a guest to make it simple to troubleshoot but I cant find on syslog or messages who is the DHCP that is assigning that IPv6 after being deleted . by the way it is global scope and dynamic and managmtaddr or something like that. with a lifetime like 3598 seconds IIRC.
I have 2 questions : on /var/log/dhcp there are no files at all and no dhclient.leases file like usually there is on Debian I think ... not sure, so where is that file or place to find who is the DHCP assigning that IPv6? Although my host PVE neither the PBS have any DHCP server running BTW , no dhcpd or anything similar. they only run the dhcp client i guess likes is normally done.
The only DHCP that I actually remember testing is on a Solaris VM I have but has been and it is off for like a year now , could it be that even if that solaris is OFF maybe that IPv6 was originally assign by that Solaris and because is off it keeps being used?, that makes no sense but ... I cant find logic any more.
Also grep dhcp messages or syslog does not show any dhcp leases? or I cant find them. I have not turn off or reboot all including the router though , maybe I should and I will now.
My second question is:
Is there any FLUSH command like in Solaris that will delete an IP permanently from an interface? I have been using only ip addr delete
Sorry for the long post and Thank you in advance for any help that will be appreciated.
Cheers
JJ
I decided to get rid of my IPv6 or Dual Stack setup so I deleted everything from cisco RV-325 router, Proxmox PVE host and all LxC and VM (PBS is a virtual machine BTW. All of them are ok but except 2 guests and now I see the same behaveour on the PVE host.
Issue is: I delete the IPv6 addr with ip addr del 2001:blabla/64 dev eth0 or ens0p18 and it gets deleted but like 5 or 10 seconds later ip addr show shows it again. it happens on 1 LxC that uses vmbr1 and 1 VM that uses vmbr2 , the resto of containers and VM it does not happen.
By I deleted ipv6 everywhere I mean through the PVE GUI all the time and I have confirmed that on all /etc/network/interfaces there is no mention to IPv6. also the cisco RV-325 is no longer dual stack and I even deleted my HE.com IPv6 tunnel just in case.
On the host PVE itself, vmbr1 and vbmr2 are manual and empty like no IP (I use vbmr0 and 3 as the host own dedicated interface and those are static ) so all that seems ok I think. But like I said after deleting that IPv6 on even the vbmr1 and 2 they come back as well right away but it says tentatively first and then dynamic. Even named on syslog mentions that is listening on that Ipv6 yet on its bind9 config file I have commented with // the line to enable listening on v6 .
On my cisco router I have also double check is no longer dual stack, no longer has a tunnel, and it is "supposedly" the only DHCP in my network or setup. I even shutdown all containers and VM except PBS and the PVE host it self that has it as a guest to make it simple to troubleshoot but I cant find on syslog or messages who is the DHCP that is assigning that IPv6 after being deleted . by the way it is global scope and dynamic and managmtaddr or something like that. with a lifetime like 3598 seconds IIRC.
I have 2 questions : on /var/log/dhcp there are no files at all and no dhclient.leases file like usually there is on Debian I think ... not sure, so where is that file or place to find who is the DHCP assigning that IPv6? Although my host PVE neither the PBS have any DHCP server running BTW , no dhcpd or anything similar. they only run the dhcp client i guess likes is normally done.
The only DHCP that I actually remember testing is on a Solaris VM I have but has been and it is off for like a year now , could it be that even if that solaris is OFF maybe that IPv6 was originally assign by that Solaris and because is off it keeps being used?, that makes no sense but ... I cant find logic any more.
Also grep dhcp messages or syslog does not show any dhcp leases? or I cant find them. I have not turn off or reboot all including the router though , maybe I should and I will now.
My second question is:
Is there any FLUSH command like in Solaris that will delete an IP permanently from an interface? I have been using only ip addr delete
Sorry for the long post and Thank you in advance for any help that will be appreciated.
Cheers
JJ