Open Stack

Hi Erik,

Sorry to hear you are still having certificate issues.

Issue #2 is probably caused by issue #1. Since we hot-plug the tenant
network for the VIP, one of the first steps after the worker connects
to the amphora agent is finishing the required configuration of the
VIP interface inside the network namespace on the amphroa.

If I remember correctly, you are attempting to configure Octavia with
the dual CA option (which is good for non-development use).

This is what I have for notes:

[certificates] gets the following:
cert_generator = local_cert_generator
ca_certificate = server CA's "server.pem" file
ca_private_key = server CA's "server.key" file
ca_private_key_passphrase = pass phrase for ca_private_key
 [controller_worker]
 client_ca = Client CA's ca_cert file
 [haproxy_amphora]
client_cert = Client CA's client.pem file (I think with it's key
concatenated is what rm_work said the other day)
server_ca = Server CA's ca_cert file

That said, I can probably run through this and write something up next
week that is more step-by-step/detailed.

Michael

On Fri, Oct 19, 2018 at 2:31 PM Erik McCormick
<emccormick at cirrusseven.com> wrote:
>
> Apologies for cross-posting, but in the event that these might be
> worth filing as bugs, I wanted the Octavia devs to see it as well...
>
> I've been wrestling with getting Octavia up and running and have
> become stuck on two issues. I'm hoping someone has run into these
> before. My google foo has come up empty.
>
> Issue 1:
> When the Octavia controller tries to poll the amphora instance, it
> tries repeatedly and eventually fails. The error on the controller
> side is:
>
> 2018-10-19 14:17:39.181 26 ERROR
> octavia.amphorae.drivers.haproxy.rest_api_driver [-] Connection
> retries (currently set to 300) exhausted.  The amphora is unavailable.
> Reason: HTTPSConnectionPool(host='10.7.0.112', port=9443): Max retries
> exceeded with url: /0.5/plug/vip/10.250.20.15 (Caused by
> SSLError(SSLError("bad handshake: Error([('rsa routines',
> 'RSA_padding_check_PKCS1_type_1', 'invalid padding'), ('rsa routines',
> 'rsa_ossl_public_decrypt', 'padding check failed'), ('asn1 encoding
> routines', 'ASN1_item_verify', 'EVP lib'), ('SSL routines',
> 'tls_process_server_certificate', 'certificate verify
> failed')],)",),)): SSLError: HTTPSConnectionPool(host='10.7.0.112',
> port=9443): Max retries exceeded with url: /0.5/plug/vip/10.250.20.15
> (Caused by SSLError(SSLError("bad handshake: Error([('rsa routines',
> 'RSA_padding_check_PKCS1_type_1', 'invalid padding'), ('rsa routines',
> 'rsa_ossl_public_decrypt', 'padding check failed'), ('asn1 encoding
> routines', 'ASN1_item_verify', 'EVP lib'), ('SSL routines',
> 'tls_process_server_certificate', 'certificate verify
> failed')],)",),))
>
> On the amphora side I see:
> [2018-10-19 17:52:54 +0000] [1331] [DEBUG] Error processing SSL request.
> [2018-10-19 17:52:54 +0000] [1331] [DEBUG] Invalid request from
> ip=::ffff:10.7.0.40: [SSL: SSL_HANDSHAKE_FAILURE] ssl handshake
> failure (_ssl.c:1754)
>
> I've generated certificates both with the script in the Octavia git
> repo, and with the Openstack Ansible playbook. I can see that they are
> present in /etc/octavia/certs.
>
> I'm using the Kolla (Queens) containers for the control plane so I'm
> sure I've satisfied all the python library constraints.
>
> Issue 2:
> I"m not sure how it gets configured, but the tenant network interface
> (ens6) never comes up. I can spawn other instances on that network
> with no issue, and I can see that Neutron has the port attached to the
> instance. However, in the instance this is all I get:
>
> ubuntu at amphora-33e0aab3-8bc4-4fcb-bc42-b9b36afb16d4:~$ ip a
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
> group default qlen 1
>     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
> 2: ens3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc pfifo_fast
> state UP group default qlen 1000
>     link/ether fa:16:3e:30:c4:60 brd ff:ff:ff:ff:ff:ff
>     inet 10.7.0.112/16 brd 10.7.255.255 scope global ens3
>        valid_lft forever preferred_lft forever
>     inet6 fe80::f816:3eff:fe30:c460/64 scope link
>        valid_lft forever preferred_lft forever
> 3: ens6: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group
> default qlen 1000
>     link/ether fa:16:3e:89:a2:7f brd ff:ff:ff:ff:ff:ff
>
> There's no evidence of the interface anywhere else including udev rules.
>
> Any help with either or both issues would be greatly appreciated.
>
> Cheers,
> Erik
>
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev