3.0.2-51:ICMP TTL Expire is being answered by vRouter underlay IP
Affects | Status | Importance | Assigned to | Milestone | ||
---|---|---|---|---|---|---|
Juniper Openstack | Status tracked in Trunk | |||||
R3.0 |
Fix Committed
|
Medium
|
RAVI KIRAN | |||
R3.1 |
Fix Committed
|
Medium
|
RAVI KIRAN | |||
R3.2 |
Fix Committed
|
Medium
|
RAVI KIRAN | |||
Trunk |
Fix Committed
|
Medium
|
RAVI KIRAN |
Bug Description
When VM1 does a traceroute to VM2, I see that the first set of 3 UDP packets with TTL = 1 is being answered by vRouter with its own IP address instead of GW IP Address.
VM2 = 40.40.40.3
VM1 = 30.30.30.3
ifcoroot@
vhost0 Link encap:Ethernet HWaddr f8:bc:12:33:43:30
inet addr:10.204.74.4 Bcast:10.204.74.255 Mask:255.255.255.0
inet6 addr: fe80::fabc:
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:2395222 errors:0 dropped:8 overruns:0 frame:0
TX packets:1746387 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:1343437964 (1.3 GB) TX bytes:2055102562 (2.0 GB)
Please see below:
ubuntu@ubuntu:~$ traceroute 40.40.40.3
traceroute to 40.40.40.3 (40.40.40.3), 30 hops max, 60 byte packets
1 kingfisher.
2 30.30.30.4 (30.30.30.4) 5.642 ms 5.580 ms 5.563 ms
3 40.40.40.3 (40.40.40.3) 5.406 ms 8.606 ms 8.576 ms
Please note, this is a service chain scenario and both VMs and vSRX are on the same compute.
I think for a customer VM, the underlay should not be visible and hence the VM should not see the IP-Fabric details.
Please fix it such that vRouter responds back with GW IP instead of its own IP.
Greetings,
Sandeep.
information type: | Proprietary → Public |
tags: | added: vrouter |
Changed in juniperopenstack: | |
assignee: | nobody → Hari Prasad Killi (haripk) |
importance: | Undecided → Medium |
Changed in juniperopenstack: | |
assignee: | Hari Prasad Killi (haripk) → RAVI KIRAN (ravibk) |
Review in progress for https:/ /review. opencontrail. org/25363
Submitter: RAVI KIRAN (<email address hidden>)