Bug 1311985
| Summary: | force_metadata = True : VMs fail receiveing meatdata ( no interface with 169.254.169.254 ip in qdhcp namespace) | ||
|---|---|---|---|
| Product: | Red Hat OpenStack | Reporter: | Alexander Stafeyev <astafeye> |
| Component: | openstack-neutron | Assignee: | Assaf Muller <amuller> |
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Toni Freger <tfreger> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.0 (Kilo) | CC: | amuller, chrisw, ihrachys, nyechiel, psutter, srevivo |
| Target Milestone: | --- | Keywords: | Triaged, ZStream |
| Target Release: | 9.0 (Mitaka) | ||
| Hardware: | x86_64 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2018-06-26 18:08:14 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
Hi, I just tripped over this while playing with OSP7 allinone setup. Contrary to what was suggested above I think the IP address should be added to the qrouter namespace. My rationale behind this: - Metadata proxy runs in qrouter namespace, not qdhcp: # ip netns list qdhcp-8e62c61b-bbb8-4d39-aaf3-c192345bfbed qrouter-3054833f-f130-4d8f-ab3f-7ce54af27ec7 # ip netns pids qrouter-3054833f-f130-4d8f-ab3f-7ce54af27ec7 4247 # ps ax | grep 4247 4247 ? S 0:01 /usr/bin/python2 /bin/neutron-ns-metadata-proxy --pid_file=/var/lib/neutron/external/pids/3054833f-f130-4d8f-ab3f-7ce54af27ec7.pid --metadata_proxy_socket=/var/lib/neutron/metadata_proxy --router_id=3054833f-f130-4d8f-ab3f-7ce54af27ec7 --state_path=/var/lib/neutron --metadata_port=9697 --metadata_proxy_user=991 --metadata_proxy_group=988 --verbose --log-file=neutron-ns-metadata-proxy-3054833f-f130-4d8f-ab3f-7ce54af27ec7.log --log-dir=/var/log/neutron - Qrouter namespace has relevant iptables rule: # ip netns exec qrouter-3054833f-f130-4d8f-ab3f-7ce54af27ec7 iptables -t nat -nL | grep 169 REDIRECT tcp -- 0.0.0.0/0 169.254.169.254 tcp dpt:80 redir ports 9697 IIRC, I had a similar problem when playing with the same setup using packages from RDO when NetworkManager was still active. Have you already analyzed the cause of this issue and can provide further information on how this will be solved? Thanks, Phil Please note that the upstream patch [1] that was written (not by me) to fix this patch was recently merged. I was waiting until it would merge to address this bug again, so whomever picks this up again should be aware of said patch :) John. [1]: https://review.openstack.org/#/c/305615/ Patch 336872 was merged, which is a backport to Mitaka. Not available in OSP 9 yet, will be in the next rebase. |
Description of problem: [root@overcloud-controller-0 ~]# cat /etc/neutron/dhcp_agent.ini | grep metadata | grep -v "#" force_metadata = True enable_isolated_metadata = False enable_metadata_network = False [stack@undercloud ~]$ neutron net-list +--------------------------------------+----------------------------------------------------+-------------------------------------------------------+ | id | name | subnets | | d7ebddcd-9989-4068-a8d9-66381e83d1f5 | int_net | 739b813d-4863-44e3-acd5-0bf6c3aaec76 192.168.3.0/24 | +--------------------------------------+----------------------------------------------------+-------------------------------------------------------+ [root@overcloud-controller-0 ~]# ip netns exec qdhcp-d7ebddcd-9989-4068-a8d9-66381e83d1f5 ip a 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 36: tap7002581e-a4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN link/ether fa:16:3e:3b:e9:ae brd ff:ff:ff:ff:ff:ff inet 192.168.3.3/24 brd 192.168.3.255 scope global tap7002581e-a4 valid_lft forever preferred_lft forever inet6 fe80::f816:3eff:fe3b:e9ae/64 scope link valid_lft forever preferred_lft forever We should have interface on qdhcp namespace with 169.254.169.254 ip for metadata when "force_metadata = True" in /etc/neutron/dhcp-agent.ini. VMs are not receiving metadata in this scenario Version-Release number of selected component (if applicable): How reproducible: 100% Steps to Reproduce: 1. set force_metadata = True in /etc/neutron/dhcp-agent.ini and restart dhcp agent 2. Create net and subnet. 3. ip netns exec qdhcp-... ip a Actual results: No interface with 169.254.169.254 ip and VMs fail to receive metadata Expected results: we should have an interface with the ip address 169.254.169.254 so VMs will receive metadata Additio[root@overcloud-controller-0 ~]# rpm -qa | grep neutron openstack-neutron-bigswitch-lldp-2015.1.38-1.el7ost.noarch openstack-neutron-ml2-2015.1.2-9.el7ost.noarch python-neutronclient-2.4.0-2.el7ost.noarch python-neutron-2015.1.2-9.el7ost.noarch openstack-neutron-2015.1.2-9.el7ost.noarch openstack-neutron-lbaas-2015.1.2-1.el7ost.noarch python-neutron-lbaas-2015.1.2-1.el7ost.noarch openstack-neutron-common-2015.1.2-9.el7ost.noarch openstack-neutron-openvswitch-2015.1.2-9.el7ost.noarch openstack-neutron-metering-agent-2015.1.2-9.el7ost.noarch [root@overcloud-controller-0 ~]# rpm -qa | grep meta yum-metadata-parser-1.1.4-10.el7.x86_64nal info: