Foreman host's hostgroup update failed via api/hammer with Name has already been taken error

[Suresh_P]

Problem: in Foreman few of the host’s hostgroup update failed via api/hammer. While try to update via api/hammer getting the error “Name has already been taken error”, but there is no duplicate entry found in foreman.

Expected outcome: foreman api / hammer should update the hostgroup for hosts.

Foreman and Proxy versions: Foreman : 3.4.0-rc1, Proxy: 3.4.0~rc1-1

Foreman and Proxy plugin versions:

Distribution and version: Debian 11 (bullseye )

Other relevant data:
[ INFO 2022-08-25T01:51:03 HammerCLIForeman::Host::UpdateCommand] Called with options: {“option_volume_list”=>, “option_interface_list”=>, “option_id”=>852, “option_name”=>“10.200.70.81”, “option_hostgroup_title”=>“Test”, “option_hostgroup_id”=>408}
{“name”=>“10.200.70.81”, “hostgroup_id”=>408, “puppet_attributes”=>{}}
[DEBUG 2022-08-25T01:51:03 HammerCLIForeman::CommandExtensions::UpdateCommon] Called block for HammerCLIForeman::UpdateCommand request params:
#<Proc:0x00005631a99d3038 /usr/lib/ruby/vendor_ruby/hammer_cli_foreman/command_extensions/update_common.rb:8>
[DEBUG 2022-08-25T01:51:03 HammerCLIForemanPuppet::CommandExtensions::HostPuppetProxy] Called block for HammerCLIForeman::Host::UpdateCommand request params:
#<Proc:0x00005631aa0abdd8 /usr/lib/ruby/vendor_ruby/hammer_cli_foreman_puppet/command_extensions/host.rb:37>
[ INFO 2022-08-25T01:51:03 API] Server: https://foreman.example.com
[ INFO 2022-08-25T01:51:03 API] PUT /api/hosts/852
[DEBUG 2022-08-25T01:51:03 API] Params: {
“host” => {
“name” => “10.200.70.81”,
“hostgroup_id” => 408,
“puppet_attributes” => {}
}
}
[DEBUG 2022-08-25T01:51:03 API] Headers: {}
[DEBUG 2022-08-25T01:51:03 API] Using authenticator: HammerCLIForeman::Api::InteractiveBasicAuth
[ERROR 2022-08-25T01:51:03 API] 422 Unprocessable Entity
[DEBUG 2022-08-25T01:51:03 API] {
“error” => {
“id” => 852,
“errors” => {
“name” => [
[0] “has already been taken”
]
},
“full_messages” => [
[0] “Name has already been taken”
]
}
}
[DEBUG 2022-08-25T01:51:03 Exception] Using exception handler HammerCLIForeman::ExceptionHandler#handle_unprocessable_entity
[ERROR 2022-08-25T01:51:03 Exception] Name has already been taken
Could not update the host:
Name has already been taken

[Suresh_P]

any one please look into it

[Suresh_P]

Hi Guys,

We found issue causing scenario in /usr/share/foreman/app/models/host/base.rb file.

Hence we disabled “:uniqueness => true” check.

Before Change:
validates :name, :presence => true, :uniqueness => true, :format => {:with => Net::Validations::HOST_REGEXP, :message => _(Net::Validations::HOST_REGEXP_ERR_MSG)}

After change:
validates :name, :presence => true, :format => {:with => Net::Validations::HOST_REGEXP, :message => _(Net::Validations::HOST_REGEXP_ERR_MSG)}

But we are not sure about the impact of this change, we did this as a work around only.

Someone please check and get back to us on the actual issue.

Thanks & Regards,
Suresh

[buzzsaw.code]

This is still an issue with foreman-3.6.1 and the above work around does work but is there a better fix for this ?

In our scenario, we were updating just the device owner in Foreman via the API for an existing device which returned the “Name has already been taken”.

I understand the need to enforce this during a name attribute update, but in this case we are only updating the owner and not the name at all.

[buzzsaw.code]

This is still around in Foreman 3.9.3, is there still no fix for this beyond turning off name uniqueness ?

[buzzsaw.code]

Issue is still around in 3.12…

[angry_yak_shaver]

Happy 2025!

After upgrading to 3.13 we are hitting this problem as well.

The problem can be reproduced from the webUI and via API calls to create a host record with a hostgroup set.

The record is created successfully but the HTTP POST call to /hosts returns the error.

Any pointers?

[angry_yak_shaver]

Turns out my problem is unrelated - we have webhooks calling shellhooks and once we disabled them it started working again.

Strangely enough even without any erb expansion and using the example debug shellhooks we are still seeing issues - will gather more data and start a new thread instead.

[cps86]

Hello

@Marek_Hulan I think there’s a bug in foreman which casues it.
It was reported on the forums many times, e.g.:

Validation failed: Name has already been taken, when registering a client

Are you really sure it doesn’t exist in foreman? Did you check with hammer? # hammer host list

Name has already been taken

Problem: I am using Foreman 2.1.0 with Puppetserver 6.16. After I run “puppet agent -t” on my agent machine, it shows up in Foreman console. Good. But then when I try to add this host to any host group or change anything, either from Foreman console or with hammer command, I got error: “Name has already been taken”. This happened to 7 out of my 300 hosts. Expected outcome: Host can be added to a host group. Foreman and Proxy versions: Foreman 2.1.0 Foreman and Proxy plugin versions: Distr…

Bootstrap host registration failure - Name has already been taken

hey @evgeni, Yes I have cycled through different organizations and locations combinations to check for an instance of the host name. Is hasn’t been fruitful.

HTTP error (422 - Unknown): Validation failed: Name has already been taken

Here is the OS list, same info in UI. # hammer os list ---|-----------------------|--------------|------- ID | TITLE | RELEASE NAME | FAMILY ---|-----------------------|--------------|------- 2 | CentOS 7 | | Redhat 3 | CentOS 7.7.1908 | | Redhat 1 | CentOS Linux 7.8.2003 | | Redhat 4 | CentOS Linux 7.9.2009 | | Redhat ---|-----------------------|--------------|------- Here are the facts for the problema…

I am using Foreman v. 3.13 and theforeman.foreman.host module (theforeman.foreman.host module – Manage Hosts — Ansible Community Documentation) with Ansible and I am experiencing the same problem. I can’t add a host to Foreman, because, allegedly the name is already taken!
So after reading this and other threads I queried the database to check if hosts is added:

select id, name, type, organization_id, location_id from hosts where name like '%ee0407%';
id|name|type|organization_id|location_id
2497|ee0407||1|9
(1 row)

It’s there! Some other host, which is really added to foreman and can see that host in the UI:

select id, name, type, organization_id, location_id from hosts where name like '%ee0408%';
id|name|type|organization_id|location_id
2495|ee0408|Host::Managed|1|9

As you can see the only difference is the type, which is set to Host::Managed.
Now, if host exists in the foreman database, can someone please explain why:

1. Host is not being displayed in the gui?
2. I can’t retrieve the host details using the foreman API, e.g:
   Http get call to: /api/hosts/2497

{
“error”: {“message”:“Resource host not found by id ‘2497’”}
}

3. Regardless if host is already to Foreman or not, ansible foreman.host module should be indempotent, right? So if host exists, should update its properties. If not, it should add it. However, it fails with:

“msg”: “Failed to ensure entity state: ForemanApiException: Error while performing create on hosts: 422 Client Error: Unprocessable Content for url: /api/hosts - {‘id’: None, ‘errors’: {‘name’: [‘has already been taken’]}, ‘full_messages’: [‘Name has already been taken’]}”

ID: None, in the result which is weird.

4. Adding state: absent to the ansible module doesn’t work:

"changed": false,
"diff": {
    "after": {
        "hosts": [
            {}
        ]
    },
    "before": {
        "hosts": [
            {}
        ]
    }
},
"entity": {
    "hosts": [
        null
    ]
},

[cps86]

Personally, I think it’s caused by Foreman Ansible Callback plugin.
When you can’t add a host to foreman by whatever reason, callback plugin still reports that host and it will be added to Foreman automatically if this option is turned on (which is default).

Perhaps it reports it as a non managed type?

[siva]

Does this change fix the duplication issue @cps86 ?