Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
This is now the job of the certmonger_user profile. So these bits are
not needed anymore in the service profiles.
Change-Id: Iaa3137d7d13d5e707f587d3905a5a32598c08800
Depends-On: Ibf58dfd7d783090e927de6629e487f968f7e05b6
|
|
The rabbitmq user check is moved to step >= 2 from step >= 1. There is
no gaurantee that rabbitmq is running at step 1, especially if updating
a failed stack that never made it past step 1 to begin with.
Change-Id: I029193da4c180deff3ab516bc8dc2da14c279317
Closes-Bug: #1675194
|
|
This optionally enables TLS for RabbitMQ in the internal network. Note
that this leaves enable_internal_tls as undef instead of using the
regular default. This is because we don't want to enable this just now,
since we first want to pass the necessary hieradata via t-h-t. This will
be cleaned in further commits.
bp tls-via-certmonger
Depends-On: I4f37e77ae12e9582fab7d326ebd4c70127c5445f
Depends-On: Ic32b2cb253fa0dc43aad7226b24919b7e588faa9
Change-Id: Ic2a7f877745a0a490ddc9315123bd1180b03c514
|
|
Rabbitmq Password is set on the fresh deployment, but during
update, if the password is changed, it is modified in all config
files including rabbitmq config. But the rabbitmq connection fails
because the new password is not successful applied to rabbitmq.
Setting the rabbitmq_user will invoke 'rabbitmqctl change_password'.
Scenario: The password change is applied on Step1 when configuring
Rabbitmq. Other services may be updated on different Steps. Till
other services config is updated with new rabbitmq password, and
restarted, the connections will get Access Denied response. It has
cyclic dependency. So the passwords will be changes at Step1 and
once all services are updated, the connections will work as is.
Partial-Bug: #1611704
Change-Id: I44865af3d5eb2d37eb648ac7227277e86c8fbc54
|
|
Currently the inter-cluster communication port listens to all ip
addresses:
tcp 0 0 0.0.0.0:25672 0.0.0.0:* LISTEN 25631/beam.smp
In order to limit it to listen only to the network assigned to rabbitmq
we need to add the following:
{kernel, [
...
{inet_dist_use_interface, {172,17,0,16}},
...
]}
In order to do the conversion from an ip address to the Erlang
representation we add a function that takes a string and returns a
converted output. The (~400 randomly generated) IPv6/4 addresses at [1]
have been parsed both via erl's built-in inet:parse_address() function
and our ruby implementation. All converted ip addresses resulted in the
same output [2], [3]. The only difference is that Erlang's parse_address()
considers network ip addresses (e.g. 10.0.0.0) invalid whereas the ruby
function does not. This should not be a problem as the use case here is
to bind a service to a specific ip address on an interface and if
anything we likely prefer the less strict behaviour, given that at least
in theory it is perfectly valid for an interface to have a network
address assigned to it.
[1] http://acksyn.org/files/tripleo/ip-addresses.txt
[2] http://acksyn.org/files/tripleo/ip-addresses-ruby.txt
[3] http://acksyn.org/files/tripleo/ip-addresses-erl.txt
Change-Id: I211c75b9bab25c545bcc7f90f34edebc92bba788
Partial-Bug: #1645898
|
|
This sets the cluster_nodes configuration in RabbitMQ to use FQDNs
instead of IP addresses. Note that in HA, RabbitMQ is already
configured using FQDNs.
Change-Id: I2b1cec25ff25f4afd72a28246c2cda9c58d7b61e
|
|
In change I35921652bd84d1d6be0727051294983d4a0dde10 we want to remove
all those duplicate tcp_listen_option entries. One consequence of that
is that we need to set rabbitmq::tcp_keepalive to true via hiera
(as opposed to forcing it via the tcp_listen_option hash).
For this to work we need to remove this forced parameter override.
Note that even if I35921652bd84d1d6be0727051294983d4a0dde10 and this
change don't merge at the exact same time it is still okay because
we do force tcp_keepalive to true via the tcp_listen_options.
Change-Id: I608477d5714a5081b3b4ab3b9fc2932bdd598301
|
|
When deploying via ipv6, rabbitmq-ctl commands have the following
issues:
- `rabbitmq cluster_status` shows nodedown alerts
- list_queues / list_connections hang
- `rabbitmqctl node_health_check` fails with an error.
* There is no any issue while performing activity on RHOS setup(From
* horizon/cli). i.e. RHOS environment is functioning as expected.
For example:
sudo rabbitmqctl node_health_check -n rabbit@node1
Checking health of node 'rabbit@node1' ...
Heath check failed:
health check of node 'rabbit@node1' fails: nodedown
The problem is that we are missing the following in
/etc/rabbitmq/rabbitmq-env.conf:
RABBITMQ_CTL_ERL_ARGS="-proto_dist inet6_tcp"
Fix these by setting the appropriate RABBITMQ_CTL_ERL_ARGS when
deploying ipv6.
Closes-Bug: #1633693
Change-Id: I53f4e76e687b3966fbb74fd0c2d83f05176630de
|
|
These hiera keys aren't aligned with the service names, which
will be required for composable generation of the ip lists
per service.
Change-Id: I423b544df174254ac511b906b0c570e701678022
Depends-On: I7febf28bf409e25e8e5961ab551b6d56bb11e0c6
Partially-Implements: blueprint custom-roles
|
|
As we are staring to manually check overcloud services
the first step is to check that the puppet profiles
are all aligned.
Changes applied:
No logic added or removed in this submission.
Removed unused parameters.
Align header comments structure.
All profiles parameters sorted following:
"Mandatory params first sorted alphabetically
then optional params sorted alphabetically."
Note: Following submissions will check pacemaker,
cinder, mistral and redis services in the base profiles
as some of them has the $pacemaker_master parameter
defaulted to true.
Change-Id: I2f91c3f6baa33f74b5625789eec83233179a9655
|
|
We were not consuming the rabbitmq_* static hiera settings when
deploying without pcmk and with a single controller.
Change-Id: I1506093e3d4365e2617521737c8f53edfb022133
|
|
Add RabbitMQ composable role, and keep the same logic that we had in
THT.
Implements: blueprint refactor-puppet-manifests
Change-Id: I961bdbe1cc6dd1d4a315de616439f9fc77d793ae
|