Open Stack

Did you change the nova_metadata_ip option to nova_metadata_host in 
metadata_agent.ini?  The former value was deprecated several releases ago 
and now no longer functions as of pike.  The metadata service will throw 
500 errors if you don't change it.

On November 12, 2018 19:00:46 Ignazio Cassano <ignaziocassano at gmail.com> wrote:
> Any other suggestion ?
> It does not work.
> Nova metatada is on port 8775 in listen but no way to solve this issue.
> Thanks
> Ignazio
>
> Il giorno lun 12 nov 2018 alle ore 22:40 Slawomir Kaplonski 
> <skaplons at redhat.com> ha scritto:
> Hi,
>
> From logs which You attached it looks that Your neutron-metadata-agent 
> can’t connect to nova-api service. Please check if nova-metadata-api is 
> reachable from node where Your neutron-metadata-agent is running.
>
>> Wiadomość napisana przez Ignazio Cassano <ignaziocassano at gmail.com> w dniu 
>> 12.11.2018, o godz. 22:34:
>>
>> Hello again,
>> I have another installation of ocata .
>> On ocata the metadata for a network id is displayed by ps -afe like this:
>>  /usr/bin/python2 /bin/neutron-ns-metadata-proxy 
>>  --pid_file=/var/lib/neutron/external/pids/c4731392-9b91-4663-adb3-b10b5ebcc4f1.pid 
>>  --metadata_proxy_socket=/var/lib/neutron/metadata_proxy 
>>  --network_id=c4731392-9b91-4663-adb3-b10b5ebcc4f1 
>>  --state_path=/var/lib/neutron --metadata_port=80 --metadata_proxy_user=996 
>>  --metadata_proxy_group=993 
>>  --log-file=neutron-ns-metadata-proxy-c4731392-9b91-4663-adb3-b10b5ebcc4f1.log 
>>  --log-dir=/var/log/neutron
>>
>> On queens like this:
>>  haproxy -f 
>>  /var/lib/neutron/ns-metadata-proxy/e8ba8c09-a7dc-4a22-876e-b8d4187a23fe.conf
>>
>> Is it the correct behaviour ?
>
> Yes, that is correct. It was changed some time ago, see 
> https://bugs.launchpad.net/neutron/+bug/1524916
>
>>
>> Regards
>> Ignazio
>>
>>
>>
>> Il giorno lun 12 nov 2018 alle ore 21:37 Slawomir Kaplonski 
>> <skaplons at redhat.com> ha scritto:
>> Hi,
>>
>> Can You share logs from Your haproxy-metadata-proxy service which is 
>> running in qdhcp namespace? There should be some info about reason of those 
>> errors 500.
>>
>> > Wiadomość napisana przez Ignazio Cassano <ignaziocassano at gmail.com> w 
>> dniu 12.11.2018, o godz. 19:49:
>> >
>> > Hi All,
>> > I upgraded  manually my centos 7 openstack ocata to pike.
>> > All worked fine.
>> > Then I upgraded from pike to Queens and instances stopped to reach 
>> metadata on 169.254.169.254 with error 500.
>> > I am using isolated metadata true in my dhcp conf and in dhcp namespace  
>> the port 80 is in listen.
>> > Please, anyone can help me?
>> > Regards
>> > Ignazio
>> >
>> > _______________________________________________
>> > OpenStack-operators mailing list
>> > OpenStack-operators at lists.openstack.org
>> > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>>
>> —
>> Slawek Kaplonski
>> Senior software engineer
>> Red Hat
>>
>
> —
> Slawek Kaplonski
> Senior software engineer
> Red Hat
>
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20181112/09c1a950/attachment.html>