Ask Your Question

Mrxlazuardin's profile - activity

2020-02-19 05:57:51 -0500 received badge  Famous Question (source)
2020-02-19 05:57:51 -0500 received badge  Notable Question (source)
2019-12-14 08:33:38 -0500 received badge  Notable Question (source)
2019-12-02 00:48:22 -0500 received badge  Popular Question (source)
2019-12-01 09:36:44 -0500 asked a question OSA with Different Deployment Address

Hi,

I'm looking the way about how to customize the deployment where the deployment host cannot reach management interface of deployed hosts. I have setup following config.

  <deployed hostname>:
    ip: <deployed host interface address which can be reached by deployment host>
    container_vars:
      container_address: <deployed host management interface address>

After running setup-host.yaml and setup-infrastructure.yml playbooks, it seem that the deployment still use "ip" value than "container_address" value for infrastructure communication addresses which are not management interface addresses. How should I solve this case?

Best regards,

2019-09-12 00:07:44 -0500 received badge  Taxonomist
2019-08-22 06:38:20 -0500 received badge  Famous Question (source)
2019-08-14 08:39:13 -0500 received badge  Famous Question (source)
2019-02-28 03:46:55 -0500 received badge  Famous Question (source)
2019-02-28 03:46:55 -0500 received badge  Notable Question (source)
2018-10-16 03:43:01 -0500 received badge  Famous Question (source)
2018-10-07 12:36:44 -0500 received badge  Popular Question (source)
2018-10-06 16:22:12 -0500 received badge  Commentator
2018-10-06 16:22:12 -0500 commented answer Different Snapshot Time, Persistent vs Ephemeral Booted

Hi Bernd,

I just following http://docs.ceph.com/docs/master/rbd/.... How to make that ephemeral storage use RBD snapshot for doing instance snapshot like on persistent storage (volume) and use RBD snapshot as new item on image store without having to do long time copying.

Best regard

2018-10-03 14:28:59 -0500 asked a question Different Snapshot Time, Persistent vs Ephemeral Booted

Hi,

On our OpenStack + Ceph infrastructure, we find that the snapshot time between VMs which are booted from volume (persistent) are so fast due to it is just to do Ceph snapshot. But, for VMs which are booted from ephemeral are so long, they take some minutes, depends on the size of ephemeral disk. As long as I know, ephemeral booted VMs are snapshots from base images on Ceph. Why those kinds of VM have different snapshot time? Is it about different method of snapshot? Why the ephemeral booted VMs take uploading time where persistent booted VMs don't do that? It seem that ephemeral booted VMs snapshot is done on compute nodes but persistent booted VMs snapshot is done on Ceph itself, isn't it?

Best regards,

2018-08-22 08:32:28 -0500 received badge  Notable Question (source)
2018-08-21 08:34:07 -0500 received badge  Notable Question (source)
2018-08-17 09:05:29 -0500 received badge  Popular Question (source)
2018-08-14 05:09:46 -0500 commented question Horizon: Domain Admin Cannot Create User

Hi Bernd,

There in no Create button on Users menu of Horizon if I login with domain admin.

Best regards,

2018-08-14 00:25:32 -0500 asked a question Horizon: Domain Admin Cannot Create User

Hi,

I have setup multi domain environment and do delegation to domain admin. Domain admin can see all things inside its own domain and can manage projects inside its own domain. But, domain admin cannot create users inside its own domain. It seem something wrong with "rule:admin_and_matching_user_domain_id" which is related to "user.domain_id". What is happen on my case?

Best regards,

2018-08-14 00:18:51 -0500 commented answer Multidomain Horizon Dashboard

Hi Bernd,

Suddenly, I have solved this problem. I don't know how but I just open related stanza on local_settings even one with default content. Maybe there is some default which os different with the assumption.

Best regards,

2018-08-12 13:36:16 -0500 commented answer Multidomain Horizon Dashboard

Hi Bernd,

It's still not working even I have toggled domains_panel to be True on ANGULAR_FEATURES stanza.I still cannot see domain management related view. Any working step by step example I can refer for this?

Best regards,

2018-08-12 13:33:52 -0500 received badge  Popular Question (source)
2018-08-10 06:37:09 -0500 received badge  Notable Question (source)
2018-08-09 13:34:20 -0500 asked a question Multidomain Horizon Dashboard

Hi,

I have used https://github.com/openstack/keystone... on /etc/openstack-dashboard/keystone_policy.json and /etc/keystone/policy.json and use 'default' for admin_domain_id. I still cannot show domain administration panel on Horizon even I have logged in as admin (from previously single domain installation). What have I missed?

Best regards,

2018-08-09 12:20:42 -0500 commented answer Project Admin

Hi Bernd,

I have used policy.v3cloudsample.json for /etc/openstack-dashboard/keystone_policy.json and /etc/keystone/policy.json. But I still cannot see domains panel on Horizon. What have I missed?

Best regards,

2018-08-08 09:50:02 -0500 commented answer Project Admin

Hi Bernd,

If I use multi domain, what should I do for modifying existing single domain OpenStack to be multidomain OpenStack? Can we do quota allocation like on project quota on delegated domain?

Best regards,

2018-08-07 15:10:42 -0500 received badge  Popular Question (source)
2018-08-06 14:34:44 -0500 asked a question Project Admin

Hi,

I have put following update on /etc/openstack-dashboard/keystone_policy.json and /etc/keystone/policy.json.

{
    "project_admin": "role:project-admin and project_id:%(target.project.id)s",
    "identity:get_user": "rule:admin_or_owner or rule:project_admin",
    "identity:list_users": "rule:admin_required or rule:project_admin",
    "identity:create_user": "rule:admin_required or rule:project_admin",
    "identity:update_user": "rule:admin_required or rule:project_admin",
    "identity:delete_user": "rule:admin_required or rule:project_admin"
}

But the user with project-admin role still cannot see his own project users and cannot create other users for that project. What have I missed?

Best regards,

2018-07-31 02:27:55 -0500 received badge  Popular Question (source)
2018-07-29 15:30:48 -0500 asked a question openstack-keystone Systemd unit file on Queen

Hi,

I'm trying to install HA of Queen version on CentOS 7 but I cannot find openstack-keystone Systemd unit file, but Keystone is running after I run httpd. How can I set pcs resource regarding to https://docs.openstack.org/ha-guide/c... It seem that identity service cannot be load-balanced, but must be failovered.

Best regards,

2018-07-02 09:58:15 -0500 received badge  Notable Question (source)
2018-07-01 01:19:42 -0500 received badge  Popular Question (source)
2018-06-28 15:15:54 -0500 asked a question OVS Offloading with Bonding

Hi,

I'm interested with OVS offloading on https://docs.openstack.org/neutron/qu.... It seem that the OVS use SR-IOV. what is the purpose of binding-profile on that docs? Is live migration still working on that way? How to do ovs offload like that but with bonding of dual port ConnectX-5 EN NIC (ex. enp3s0f0 and enp3s0f1)? For other kernel networking traffic (management, API, Ceph, etc.), should I use PF or VF?

Best regards,

2018-06-07 02:14:37 -0500 received badge  Popular Question (source)
2018-06-07 02:14:37 -0500 received badge  Notable Question (source)
2018-06-06 05:47:20 -0500 commented answer L3 Agent Without Network Nodes

I'm thinking to move L3 agent or likely instance to Compute nodes like moving of LBaaS to LBaaSv2 (Octavia) for scalability. Any thought?

2018-06-05 21:18:45 -0500 commented question Ceph multiattach support

Hi,

I'm waiting for this one too since Ceph itself supports multiattach RBD image.

Best regards,

2018-06-05 21:13:40 -0500 answered a question Glance image upload uses swap

Hi Michael,

Maybe you can adjust vm.swappiness parameter on your controller.

Best regards,

2018-06-05 13:51:43 -0500 asked a question L3 Agent Without Network Nodes

Hi,

If I don't have dedicated Network node, which one is better for L3 agent and other Neutron agents (on Controller nodes or on Compute nodes) except L2 agent (it is seem mandatory on Compute node) and DHCP agent (since I have redundant Controller nodes and it only requires small number of agents, I prefer to place it on Controller node)? Any reason of that recommendation?

Best regards,

2017-12-01 07:19:49 -0500 received badge  Famous Question (source)
2017-11-17 08:03:06 -0500 received badge  Notable Question (source)
2017-11-17 08:03:06 -0500 received badge  Famous Question (source)
2017-11-14 05:32:44 -0500 received badge  Popular Question (source)
2017-11-13 17:25:52 -0500 asked a question Keystone High Availability

Hi,

By reading following URL, I find two ways of HA method for Keystone (Identity) API. by using HAProxy and by using Pacemaker.

https://docs.openstack.org/ha-guide/c...

https://docs.openstack.org/ha-guide/c...

Why Keystone is so special so it has/require these two ways? Or, do I miss something here?

Best regards,

2017-11-13 17:09:39 -0500 received badge  Famous Question (source)
2017-10-05 21:21:37 -0500 received badge  Notable Question (source)
2017-10-05 21:21:37 -0500 received badge  Famous Question (source)
2017-09-01 05:15:09 -0500 received badge  Famous Question (source)
2017-09-01 00:51:17 -0500 received badge  Notable Question (source)