[NETVIRT-1403] Ping between IPV6 and DCGW is not happening Created: 17/Aug/18 Updated: 19/Dec/19 Resolved: 19/Dec/19 |
|
| Status: | Resolved |
| Project: | netvirt |
| Component/s: | General |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | High |
| Reporter: | Nandini Savanura | Assignee: | Nandini Savanura |
| Resolution: | Cannot Reproduce | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Attachments: |
|
| Description |
|
When executing ipv6 ping from DCGW to VM and vice versa, ping is not working. Routes are showing up in quagga side, but quagga to Controller communication is not happening is what we suspect. Discussed with Karthikeyan and Sivakumar Perumalla regarding this. We need some help from VPN development team. Can any one from vpn team help us in this.?
Following investigations are done:
For DCGW side config: I have attached the configurations.
Below are the 2 loopback for ipv6 we configured in SER side: interface lo2 loopback ipv6 address 2001:db5::3:f816:3eff:fedb:1b40/128 ! interface lo3 loopback ipv6 address 2001:db5:4:f816:3eff:fedb:1b40::/128 no logging console
these entries are showing up in quagga: bgpd# show bgp ipv6 vpn all hidden BGP table version is 0, local router ID is 11.33.216.118 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal Origin codes: i - IGP, e - EGP, ? - incomplete
Network Next Hop Metric LocPrf Weight Path Route Distinguisher: as2 1:3
::ffff:192.168.113.94 0 100 0 ? Route Distinguisher: as2 1:4
::ffff:192.168.113.94 0 100 0 ?
::ffff:192.168.113.94 0 100 0 ? Displayed 3 out of 3 total prefixes
But in the ODL side, when we check in fib, these are missing: cli>display fib Enter password for user cscadm: RD Prefix NextHop Label Origin ------------------------------------------------------------------- 1:4 30.1.1.3/32 192.168.113.10 103012 l 1:4 2001:db5:0:2:f816:3eff:fef2:174a/128 192.168.113.11 103013 l 1:4 30.1.1.4/32 192.168.113.11 103015 l 1:4 30.1.1.1/32 0.0.0.0 103014 l 1:4 30.1.1.0/24 192.168.113.11 103009 c 1:4 2001:db5:0:2:f816:3eff:fe81:657b/128 192.168.113.10 103011 l 1:4 2001:db5:0:2::1/128 0.0.0.0 103010 l 1:4 4.4.4.4/32 192.168.113.94 786432 b 1:4 2001:db5:0:2::/64 192.168.113.11 103008 c
BGP Configuration from ODL side: cli>display bgp-config
Configuration Server Host 192.168.70.5 Port 7644
BGP Router AS-Number 100 Router-ID 11.33.216.118 Stale-Path-Time default F-bit ON Log-File /opt/quagga/var/log/quagga/zrpcd.init.log Log-Level debugging
Neighbors 192.168.113.94 AS-Number 100 Update-Source 11.33.216.118 Address-Families IPv4-Labeled-VPN IPv6-Labeled-VPN 192.168.113.95 AS-Number 100 Update-Source 11.33.216.118 Address-Families IPv4-Labeled-VPN
VPN Configs from ODL side: cli>display vpn-config
Displaying VPN config for all VPNs To display VPN config for a particular VPN, use the following syntax Usage:display vpn-config [-vid/--vpnid <id>]
VPN ID Tenant ID RD
Import-RTs
Export-RTs
Subnet IDs
------------------------------------------------------------------------------------
4ae8cd92-48ca-49b5-94e1-b2921a260001 6c53df3a-3456-11e5-a151-feff819cdc9f [1:4]
[1:4]
[1:4]
812556e5-d34e-4b95-878f-74c73792ba02 42230350-f70c-4d60-a049-a8e4edc47c
When we saw quagga logs, Vrf entry is getting deleted in the middle. I request vpn team to look into this on high priority.
root@cic-2:/opt/quagga/var/log/quagga# cat 0 | grep Vrf 2018/08/09 16:11:05 Informational ZRPC: addVrf(1:4) OK 2018/08/09 16:11:05 Informational ZRPC: addVrf(1:4, afi 2, safi 5) OK 2018/08/09 16:11:05 Informational ZRPC: addVrf(1:4, afi 1, safi 5) OK 2018/08/09 16:11:05 Informational ZRPC: delVrf(1:4, afi 2, safi 5) OK 2018/08/09 17:18:56 Informational ZRPC: addVrf(1:4) OK 2018/08/09 17:18:56 Informational ZRPC: addVrf(1:4, afi 1, safi 5) OK root@cic-2:/opt/quagga/var/log/quagga# |
| Comments |
| Comment by Karthikeyan Krishnan [ 20/Aug/18 ] |
|
Hi Nandini, Manually verified for both IPv4 and IPv6 data path traffic testing and it was working fine with DC-GW to VM(via ICMP) and vice versa without doing any code changes. Hence no fix is required for fixing this problem. |
| Comment by Nandini Savanura [ 17/Sep/18 ] |
|
This issue is seen in the latest build in the sanity(Build 250).
|
| Comment by Karthikeyan Krishnan [ 04/Oct/18 ] |
|
Hi Nandini, Requested to attach the following logs into JIRA ticket Required debug logs:
Thanks & Regards, Karthikeyan. |
| Comment by Nishchya Gupta [ 19/Dec/19 ] |
|
This issue cannot be reproduce with the latest master distribution. Please reopen the issue if seen again with below details/logs. Karaf logs(All 3 nodes) Data Store dumps output (elan/l3vpn/odl-fib/odl-l3vpn) Dump flows/groups output (all the compute nodes) Openstack Configuration commands (tested network topology commands) |