mrlesmithjr / ansible-mdadm Goto Github PK

View Code? Open in Web Editor NEW

59.0 7.0 40.0 42 KB

Jinja 100.00%

ansible ansible-role debian ubuntu raid

ansible-mdadm's Introduction

mrlesmithjr.mdadm

An Ansible role to install and manage mdadm raid arrays.

Requirements

Available unpartitioned disk devices

Role Variables

defaults/main.yml

Dependencies

None

Example Playbook

- hosts: all
  become: true
  vars:
  roles:
    - role: ansible-mdadm
  tasks:

License

BSD

Author Information

Larry Smith Jr.

ansible-mdadm's People

Contributors

Stargazers

Watchers

ansible-mdadm's Issues

Duplicate entries can be created in mdadm.conf when pre-existing mdraid devices are 'imported'

Describe the bug

If ansible-mdadm is used to manage mdraid devices created via another means, duplicate entries may be created in mdadm.conf. In the worst case, this may prevent a machine booting.

To Reproduce

As an example:

Install Centos using Anaconda. During the installation configure an mdraid device for use as the rootfs.
Anaconda creates mdadm.conf. Everything is fine so far. For example:

# mdadm.conf written out by anaconda
	MAILADDR root
	AUTO +imsm +1.x -all

	ARRAY /dev/md/pv00 level=raid1 num-devices=2 UUID=982e5d2d:b145f191:52abcf49:5ebffa20
	ARRAY /dev/md/boot level=raid1 num-devices=2 UUID=ec253f6b:45ef42ae:4182ed28:0cd28b9b
	ARRAY /dev/md/boot_efi level=raid1 num-devices=2 UUID=26460904:051f0259:cfc9daf0:9e0898eb
	ARRAY /dev/md124 level=raid1 num-devices=2 UUID=1ab96384:06a092f2:910f1eda:2255e814

Now add, for example /dev/md124 to ansible-mdadm configuration and run the role.
ansible-mdadm writes the output from mdadm --detail --scan to mdadm.conf. The regex doesn't match the existing entries, resulting in two entries per device. For example:

	# mdadm.conf written out by anaconda
	MAILADDR root
	AUTO +imsm +1.x -all

	ARRAY /dev/md/pv00 level=raid1 num-devices=2 UUID=982e5d2d:b145f191:52abcf49:5ebffa20
	ARRAY /dev/md/boot level=raid1 num-devices=2 UUID=ec253f6b:45ef42ae:4182ed28:0cd28b9b
	ARRAY /dev/md/boot_efi level=raid1 num-devices=2 UUID=26460904:051f0259:cfc9daf0:9e0898eb
	ARRAY /dev/md124 level=raid1 num-devices=2 UUID=1ab96384:06a092f2:910f1eda:2255e814

        # These added by ansible-mdadm:
	ARRAY /dev/md/pv00 metadata=1.2 name=localhost.localdomain:pv00 UUID=982e5d2d:b145f191:52abcf49:5ebffa20
	ARRAY /dev/md/boot metadata=1.2 name=localhost.localdomain:boot UUID=ec253f6b:45ef42ae:4182ed28:0cd28b9b
	ARRAY /dev/md/boot_efi metadata=1.0 name=localhost.localdomain:boot_efi UUID=26460904:051f0259:cfc9daf0:9e0898eb
	ARRAY /dev/md124 metadata=1.2 name=124 UUID=1ab96384:06a092f2:910f1eda:2255e814

Server may no longer boot due to duplicate entries.

** Expected behaviour **

No duplicate entries are created in mdadm.conf
For each mdraid device, if an entry already exists in mdadm.conf, it should be replaced using the format from ansible-mdadm.

I'm happy to submit a patch if we agree on the expected behaviour. It would be easy to match by UUID, rather than the whole ARRAY line. However, it could have side effects (removing level, num-devices from existing entries). Perhaps we should fail if this scenario happens, and prompt the user to take manual, remedial action instead?

Many thanks for your excellent work by-the-way.

Initramfs issue after creating array

Describe the bug
This is my first ever bug report, so please tell me if I can make it better.
This is potentially a duplicate of #21 .

While creating an array (Raid-5 in this case), the role runs correctly, including the arrays | Updating Initramfs task. The array is working correctly. However, after a reboot, the OS fails to mount the array because it was automatically renamed to md127.

Running update-initramfs -u manually and rebooting fixes the issue. Running this command after the role in Ansible seems to work too (cf. below).

To Reproduce
Steps to reproduce the behavior:

Fire up a VM with 1 disk for system and 3 disks for storage; install Debian
Run the role from this repo. As an example, I had the following play and variables:

  - name: Raid-5 configuration
    hosts: all  # one single host
    become: yes
    tasks:
      - name: Include mdadm role
        include_role:
          name: mrlesmithjr.mdadm
      # - name: Update initramfs (bis)
      #   command: "update-initramfs -u"
      #   when: array_created.changed

mdadm_arrays:
  - name: md0
    devices:
      - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_drive-scsi1
      - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_drive-scsi2
      - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_drive-scsi3
    filesystem: ext4
    level: '5'  # Raid type (0, 1, 5, 6...)
    mountpoint: /srv/md0
    state: 'present'

(When the three commented lines are un-commented, update-initramfs is run a second time, and things work as expected even after a reboot.)
3. Reboot the VM

Expected behavior
The Raid array should keep the same name and therefore should be mounted without any problem.

Actual behavior
At boot, the VM displays something like

[ TIME ] Timed out waiting for device dev-md0.device - /dev/md0.
[DEPEND] Dependency failed for srv-md0.mount - /srv/md0.
[DEPEND] Dependency failed for local-fs.target - Local File Systems.

The system boots in emergency mode and the lsblk command shows that the array is now called md127.

Desktop (please complete the following information):

Target OS: Debian 12.1
ansible --version:

ansible [core 2.15.1]
 config file = /path/to/project/ansible.cfg
 configured module search path = ['/home/louis/.ansible/plugins/modules', '/usr/share/ansible/plugins/modules']
 ansible python module location = /home/louis/.local/lib/python3.10/site-packages/ansible
 ansible collection location = /home/louis/.ansible/collections:/usr/share/ansible/collections
 executable location = /home/louis/.local/bin/ansible
 python version = 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] (/usr/bin/python3)
 jinja version = 3.0.3
 libyaml = True

Additional context
Note that when running the playbook, the arrays | Updating Initramfs task from this role seems to work (std_out is the usual "Generating /boot/..." and std_err is empty). But somehow, it needs to be run a second time to actually work.

Device /dev/md2 excluded by a filter.

I tried to create a RAID1 partition for LVM on an partially configured system and got the error message above.

The playbook looks like this:

---
- name: Setup myserver
  hosts: myserver
  become: true
  vars:
    mdadm_arrays:
      - name: 'md2'
        devices:
          - '/dev/sda4'
          - '/dev/sdb4'
        filesystem: 'lvm'
        level: 1
        state: present
  roles:
    - role: mrlesmithjr.mdadm

At the console of the server I saw that there were partitions on the newly made /dev/md2:

# fdisk -l /dev/md2
Disk /dev/md2: 907.7 GiB, 973951991808 bytes, 1902249984 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: dos
Disk identifier: 0x0008d9b9

Device     Boot     Start       End   Sectors  Size Id Type
/dev/md2p1           2048 139460607 139458560 66.5G 83 Linux
/dev/md2p2      139462654 142604287   3141634  1.5G  5 Extended
/dev/md2p5      139462656 142604287   3141632  1.5G 82 Linux swap / Solaris

Partition 2 does not start on physical sector boundary.

Using the advice given at https://www.simplstor.com/index.php/support/support-faqs/118-lvm-dev-excluded-by-filter I was able to create the physical volume with these commands:

# wipefs -a /dev/md2
# pvcreate /dev/md2

Now the playbook runs fine but the manual intervention contradicts the purpose of Ansible.
Can you tell me what happened here?

mdadm_conf not defined

Hello,

This is the configuration I've used based on the example from the readme.md.

    mdadm_arrays:
      - name: 'md0'
    devices:
      - '/dev/sdb'
      - '/dev/sdc'
      - '/dev/sdd'
    level: '0'
    mountpoint: '/mnt/md0'
    state: 'present'

This results in an error:

TASK [ansible-mdadm : arrays | Ensure {{ mdadm_conf }} file exists] **********************************************************************************************************************************************************************************************
Sunday 28 July 2019  17:58:57 +0200 (0:00:00.030)       0:00:07.929 *********** 
fatal: [nl-t1-kvm01]: FAILED! => {"msg": "The task includes an option with an undefined variable. The error was: 'mdadm_conf' is undefined\n\nThe error appears to be in '/Users/nan03/projects/genesis/roles/ansible-mdadm/tasks/arrays.yml': line 115, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name: arrays | Ensure {{ mdadm_conf }} file exists\n  ^ here\nWe could be wrong, but this one looks like it might be an issue with\nmissing quotes. Always quote template expression brackets when they\nstart a value. For instance:\n\n    with_items:\n      - {{ foo }}\n\nShould be written as:\n\n    with_items:\n      - \"{{ foo }}\"\n"}

This error is easily resolved by adding a variable like this:

mdadm_conf: '/etc/mdadm/mdadm.conf'

I would suggest that this variable needs a default or must be defined mandatory as part of the usage instruction, whatever you think is best.

Raid level 1 and specifing chunk size no longer work together

When trying create new RAID1 array, i got error from mdadm :
...
roles/mrlesmithjr.mdadm/tasks/arrays.yml:13

"stderr": "mdadm: specifying chunk size is forbidden for this level"

Additional info:
https://bugzilla.redhat.com/show_bug.cgi?id=1989958

man mdadm:

This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.

Differences in Centos/RHEL not correctly implimented.

Description
Multiple issues with differences between Debian and Fedora based systems. The biggest problems being the initramfs updates and the config file location.

To Reproduce

Attempt to use on Centos or other Fedora based OS

Expected behaviour
Script creates /etc/mdadm/ directory, then also creates the config file at /etc/mdadm/mdadm.conf this location is only valid on debian based systems, as such the raid will not be correctly assembled at boot.
The update-initramfs command is only valid in Debian based systems, Centos, RHEL, and Fedora all use dracut to manage their initramfs images and versioning.

It is not recommended to update the initramfs in the most common use-cases, namely when the OS is not installed on the RAID. This is actually detrimental if the raid becomes degraded, is absent at boot, or the boot drive is encrypted.

DNF is not being called as the package manager for the systems that use it.

Desktop:

OS: CentOS 7
Version: 3.10.0-1160.76.1.el7.x86_64

mdadm.conf not generated for existing array

Describe the bug

I'm using this role to configure software raid on my secondary storage. My workflow for re-provisioning is to wipe and reimage the primary storage, but leave the secondary storage intact. I then re-run the ansible against the node. I've noticed that after re-provisioning in this fashion, mdadm.conf does not get populated.

Expected behavior
mdadm.conf should be populated with arrays defined by mdadm_arrays

Additional context
I think the issue is because this task relies on the previous one reporting status changed:

ansible-mdadm/tasks/arrays.yml

Lines 121 to 129 in 104bb82

    
           # Updating mdadm.conf in order to persist between reboots 
        
           - name: arrays | Updating {{ mdadm_conf }} 
        
             lineinfile: 
        
               dest: "{{ mdadm_conf }}" 
        
               regexp: "^{{ item }}" 
        
               line: "{{ item }}" 
        
               state: "present" 
        
             with_items: '{{ array_details.stdout_lines }}' 
        
             when: array_created.changed

Tries to create a default array if a host has no `mdadm_array` variable defined

Describe the bug
When you have a host in a group that has no mdadm_arrays variable set, the role tries to create a default array.

To Reproduce
Steps to reproduce the behavior:

Add a host to an inventory that does not have /dev/sda & /dev/sdb
Do not define mdadm_arrays variable
run a playbook that include_role or depends on the role

Expected behavior
I expect the playbook to do nothing if mdadm_arrays is not defined.

Screenshots

TASK [mrlesmithjr.mdadm : arrays | Creating Array(s)] *************************************************
failed: [no_raid.test] (item={'name': 'md0', 'devices': ['/dev/sdb', '/dev/sdc'], 'filesystem': 'ext4', 'level': '1', 'mountpoint': '/mnt/md0', 'state': 'present', 'opts': 'noatime'}) => {"ansible_loop_var": "item", "changed": true, "cmd": "yes | mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sdb /dev/sdc", "delta": "0:00:00.003582", "end": "2022-05-05 23:20:30.929538", "item": {"devices": ["/dev/sdb", "/dev/sdc"], "filesystem": "ext4", "level": "1", "mountpoint": "/mnt/md0", "name": "md0", "opts": "noatime", "state": "present"}, "msg": "non-zero return code", "rc": 2, "start": "2022-05-05 23:20:30.925956", "stderr": "mdadm: cannot open /dev/sdb: No such file or directory", "stderr_lines": ["mdadm: cannot open /dev/sdb: No such file or directory"], "stdout": "", "stdout_lines": []}

Desktop (please complete the following information):

OS: Rocky 8.5

cannot create more than 1 md device, found error in task

Describe the bug
cannot create more than 1 md device, found error in task

To Reproduce
Steps to reproduce the behavior:

md3 device created on host before
adding md4 device into template

mdadm_arrays:
  - name: 'md3'
    devices:
      - '/dev/sdc'
      - '/dev/sdd'
      - '/dev/sde'
      - '/dev/sdf'
    filesystem: 'ext4'
    level: '10'
    mountpoint: '/mnt/md3'
    state: 'present'
  - name: 'md4'
    devices:
      - '/dev/sdg'
      - '/dev/sdh'
      - '/dev/sdi'
      - '/dev/sdj'
    filesystem: 'ext4'
    level: '10'
    mountpoint: '/mnt/md4'
    state: 'present'

playing you role and task no creating array:

TASK [mrlesmithjr.mdadm : arrays | Checking Status Of Array(s)]

ok: [db] => (item={u'state': u'present', u'name': u'md3', u'level': u'10', u'mountpoint': u'/mnt/md3', u'filesystem': u'ext4', u'devices': [u'/dev/sdc', u'/dev/sdd', u'/dev/sde', u'/dev/sdf']}) => {
...
    "rc": 0,
STDOUT:

md3 : active raid10 sdd[3] sdc[2] sda[0] sdb[1]

ok: [db] => (item={u'state': u'present', u'name': u'md4', u'level': u'10', u'mountpoint': u'/mnt/md4', u'filesystem': u'ext4', u'devices': [u'/dev/sdg', u'/dev/sdh', u'/dev/sdi', u'/dev/sdj']}) => {
...
    "rc": 1,
MSG:

non-zero return code

See error

md4 device not found, RC=1
But task "Creating Array(s)" skipped creatinf md4 device

TASK [mrlesmithjr.mdadm : arrays | Creating Array(s)]

skipping: [db] => (item={u'state': u'present', u'name': u'md3', u'level': u'10', u'mountpoint': u'/mnt/md3', u'filesystem': u'ext4', u'devices': [u'/dev/sdc', u'/dev/sdd', u'/dev/sde', u'/dev/sdf']})  => {
    "skip_reason": "Conditional result was False"

skipping: [db1] => (item={u'state': u'present', u'name': u'md4', u'level': u'10', u'mountpoint': u'/mnt/md4', u'filesystem': u'ext4', u'devices': [u'/dev/sdg', u'/dev/sdh', u'/dev/sdi', u'/dev/sdj']})  => {
    "skip_reason": "Conditional result was False"

Error found into file tasks/arrays.yml
It was

# Creating raid arrays
# We pass yes in order to accept any questions prompted for yes|no
- name: arrays | Creating Array(s)
  shell: "yes | mdadm --create /dev/{{ item.name }} --level={{ item.level }} --raid-devices={{ item.devices|count }} {{ item.devices| join (' ') }}"
  register: "array_created"
  with_items: '{{ mdadm_arrays }}'
  when: >
        item.state|lower == "present" and
        array_check.results[0].rc != 0

**Error found in line " array_check.results[0].rc != 0" for result index **

FIX
file tasks/arrays.yml

# Creating raid arrays
# We pass yes in order to accept any questions prompted for yes|no
- name: arrays | Creating Array(s)
  shell: "yes | mdadm --create /dev/{{ item.name }} --level={{ item.level }} --raid-devices={{ item.devices|count }} {{ item.devices| join (' ') }}"
  register: "array_created"
#  with_items: '{{ mdadm_arrays }}'
  loop: "{{ mdadm_arrays }}"
  loop_control:
    index_var: index
  when: >
        item.state|lower == "present" and
        array_check.results[index].rc != 0

I am applying index of array mdadm_arrays for array_check.results.RC

Playing after FIX

TASK [mrlesmithjr.mdadm : arrays | Creating Array(s)] *******************************************************************
skipping: [db] => (item={u'state': u'present', u'name': u'md3', u'level': u'10', u'mountpoint': u'/mnt/md3', u'filesystem': u'ext4', u'devices': [u'/dev/sdc', u'/dev/sdd', u'/dev/sde', u'/dev/sdf']})
changed: [db] => (item={u'state': u'present', u'name': u'md4', u'level': u'10', u'mountpoint': u'/mnt/md4', u'filesystem': u'ext4', u'devices': [u'/dev/sdg', u'/dev/sdh', u'/dev/sdi', u'/dev/sdj']})

TASK [mrlesmithjr.mdadm : arrays | Updating Initramfs] ******************************************************************
changed: [db]

TASK [mrlesmithjr.mdadm : arrays | Capturing Array Details] *************************************************************
ok: [db]

TASK [mrlesmithjr.mdadm : arrays | Creating Array(s) Filesystem] ********************************************************
ok: [db] => (item={u'state': u'present', u'name': u'md3', u'level': u'10', u'mountpoint': u'/mnt/md3', u'filesystem': u'ext4', u'devices': [u'/dev/sdc', u'/dev/sdd', u'/dev/sde', u'/dev/sdf']})
changed: [db] => (item={u'state': u'present', u'name': u'md4', u'level': u'10', u'mountpoint': u'/mnt/md4', u'filesystem': u'ext4', u'devices': [u'/dev/sdg', u'/dev/sdh', u'/dev/sdi', u'/dev/sdj']})

TASK [mrlesmithjr.mdadm : arrays | Mounting Array(s)] *******************************************************************
ok: [db] => (item={u'state': u'present', u'name': u'md3', u'level': u'10', u'mountpoint': u'/mnt/md3', u'filesystem': u'ext4', u'devices': [u'/dev/sdc', u'/dev/sdd', u'/dev/sde', u'/dev/sdf']})
changed: [db] => (item={u'state': u'present', u'name': u'md4', u'level': u'10', u'mountpoint': u'/mnt/md4', u'filesystem': u'ext4', u'devices': [u'/dev/sdg', u'/dev/sdh', u'/dev/sdi', u'/dev/sdj']})

on host

# cat /proc/mdstat
Personalities : [raid10] [raid1] [linear] [multipath] [raid0] [raid6] [raid5] [raid4]
md4 : active raid10 sdj[3] sdi[2] sdh[1] sdg[0]
      15000268800 blocks super 1.2 512K chunks 2 near-copies [8/8] [UUUUUUUU]
      [>....................]  resync =  0.2% (44098944/15000268800) finish=1205.9min speed=206695K/sec
      bitmap: 112/112 pages [448KB], 65536KB chunk

Additional options for raid members

Is your feature request related to a problem? Please describe.
There are some cases when devices require additional options, f.e. --write-mostly (when disks have asymmetric size or different base disk types, ref. https://raid.wiki.kernel.org/index.php/Write-mostly)

Describe the solution you'd like
Additional flag/options can be passed to device members

Describe alternatives you've considered
No alternative solution yet

Additional context
No additional context

chunk size configuration

Currently, there is no way to configure chunk size or send opts to mdadm command.

new release?

Is there any chance that current version gets released any time soon?

I'm using Ansible galaxy for managing dependencies, and it seem last release accessible thru there was from 2 years ago https://galaxy.ansible.com/ui/standalone/roles/mrlesmithjr/mdadm/ .

P.S. Thank you for your (and contributors) hard work

[Question] Is there anyway to specify custom options during array creation ?

Hi there,

Is there any way to specify options during array creation, f.e. I wanna specify --write-mostly to one of member of raid 1 ?

`update-initramfs -u` is missing

Describe the bug

Seems like final step is missing: update-initramfs -u

# mdadm.conf
#
# !NB! Run update-initramfs -u after updating this file.
# !NB! This will ensure that initramfs has an uptodate copy.

To Reproduce
Steps to reproduce the behavior:

Buy server at hetzner auction (there will be no cost, since you can cancel server after experiment, that will not take too much time) with 2 disks without hardware raid
Ask to add another 2 drives
Use installimage (comment out 3rd and 4th disks to setup 2nd raid array after installation using this ansible role)
Use this role to build 2nd raid using following playbook

- hosts: all
  become: true
  vars:
    mdadm_arrays:
    - name: 'md3'
      devices:
        - '/dev/sdc'
        - '/dev/sdd'
      filesystem: 'ext4'
      level: '1'
      mountpoint: '/data'
      state: 'present'
      opts: 'noatime'
  roles:
    - role: mrlesmithjr.mdadm
  tasks:

Reboot server

Expected behavior
Server boots.

Actual behavior
Server doesn't boot.

Desktop (please complete the following information):

Target OS: Ubuntu 20.04
Version

ansible [core 2.12.1]
  config file = None
  configured module search path = ['~/.ansible/plugins/modules', '/usr/share/ansible/plugins/modules']
  ansible python module location = /usr/local/Cellar/ansible/5.1.0/libexec/lib/python3.10/site-packages/ansible
  ansible collection location = ~/.ansible/collections:/usr/share/ansible/collections
  executable location = /usr/local/bin/ansible
  python version = 3.10.1 (main, Dec  6 2021, 23:19:43) [Clang 12.0.0 (clang-1200.0.32.29)]
  jinja version = 3.0.3
  libyaml = True

~/.ansible/roles/mrlesmithjr.mdadm/meta/.galaxy_install_info:version: v0.1.1

Task fails at Removing arrays

Describe the bug
Task fails when trying to destroy array by setting state=absent

To Reproduce
Steps to reproduce the behavior:

Change state: present to state: absent in playbook

 vars:
    mdadm_arrays:
      - name: 'md0'
        devices:
          - '/dev/sdb'
          - '/dev/sdc'
        filesystem: 'lvm'
        level: 1
        state: present

Output from play:

TASK [ansible-mdadm : arrays | Removing Array(s)] **************************************************************************************************************************
failed: [storage.xyz.abcdef.com] (item={'name': 'md0', 'devices': ['/dev/sdb', '/dev/sdc'], 'filesystem': 'lvm', 'level': 1, 'state': 'absent'}) => changed=true 
  ansible_loop_var: item
  cmd:
  - mdadm
  - --remove
  - /dev/md0
  delta: '0:00:00.002178'
  end: '2022-02-04 12:48:51.857826'
  item:
    devices:
    - /dev/sdb
    - /dev/sdc
    filesystem: lvm
    level: 1
    name: md0
    state: absent
  msg: non-zero return code
  rc: 1
  start: '2022-02-04 12:48:51.855648'
  stderr: 'mdadm: error opening /dev/md0: No such file or directory'
  stderr_lines: <omitted>
  stdout: ''
  stdout_lines: <omitted>

Expected behavior
The playbook should complete without errors

Desktop (please complete the following information):

OS: RedHat Enterprise Linux 8.5
mdadm: mdadm-4.2-rc2.el8.x86_64

Additional context
It seems like the MD device is already removed from the system from the task before, "Stopping raid arrays in preparation of destroying". I don't know if the device is removed on systems other than RHEL

mdadm.conf entries do not use expected array name

Describe the bug

On Rocky 9, the mdadm.conf is being generated with lines similar to:

ARRAY /dev/md/cpt01:0 metadata=1.2 name=cpt01:0 UUID=1c22a1b7:83aea9e8:32ddbd4a:1ff7d63c
ARRAY /dev/md/localhost.localdomain:0 metadata=1.2 name=localhost.localdomain:0

I am setting the name to be md127 in mdadm_arrays.

To Reproduce

Run role against rocky 9 host
Check /etc/mdadm.conf

Expected behavior

I would expect the config file to look more like:

ARRAY /dev/md/127 metadata=1.2 name=cpt01:0 UUID=1c22a1b7:83aea9e8:32ddbd4a:1ff7d63c
ARRAY /dev/md/localhost.localdomain:0 metadata=1.2 name=localhost.localdomain:0

** Additional context **

There is some context in on in this issue in this bug report: https://bugzilla.redhat.com/show_bug.cgi?id=606481#c1. In particular this comment:

For version 1.2 superblocks, the preferred way to create arrays is by using a name instead of a number.  For example, if the array is your home partition, then creating the array with the option --name=home will cause the array to be assembled with a random device number (which is what you are seeing now, when an array doesn't have an assigned number we start at 127 and count backwards), but there will be a symlink in /dev/md/ that points to whatever number was used to assemble the array.  The symlink in /dev/md/ will be whatever is in the name field of the superblock.  So in this example, you would have /dev/md/home that would point to /dev/md127 and the preferred method of use would be to access the device via the /dev/md/home entry.

I found that using the scan with the name set in the array generated it correctly:

- hosts: compute-regular-gen2
  gather_facts: false
  tasks:
    # Capture the raid array details to append to mdadm.conf
    # in order to persist between reboots
    - name: arrays | Capturing Array Details
      command: "mdadm --detail --scan /dev/{{ item.name }}"
      register: "array_details"
      changed_when: false
      with_items: "{{ mdadm_arrays }}"
      become: true
    
    - name: arrays | Updating /etc/mdadm.conf
      lineinfile:
        dest: /etc/mdadm.conf
        regexp: "^{{ item }}"
        line: "{{ item }}"
        state: "present"
      with_items: "{{ array_details.results | map(attribute='stdout_lines') | flatten }}"
      become: true

I think that needs work as essentially we want to match on the UUID to replace an existing entry, but putting it out there to solicit some feedback.

Support names that don't match md*

I'm wondering if we could support passing --name to create. Interested to here what you think and if we could do that in a backwards compatible way to support something like:

mdadm_arrays:
 - name: 'secondary'
   # Define disk devices to assign to array
   devices:
     - '/dev/sdb'
     - '/dev/sdc'
   # Define filesystem to partition array with
   filesystem: 'ext4'
   # Define the array raid level
   # 0|1|4|5|6|10
   level: '1'
   # Define mountpoint for array device
   mountpoint: '/mnt/md0'
   # Define if array should be present or absent
   state: 'present'

From reading: https://bugzilla.redhat.com/show_bug.cgi?id=606481#c1, this is seems to the recommend way to address a particular raid array.

	# Updating mdadm.conf in order to persist between reboots
	- name: arrays \| Updating {{ mdadm_conf }}
	lineinfile:
	dest: "{{ mdadm_conf }}"
	regexp: "^{{ item }}"
	line: "{{ item }}"
	state: "present"
	with_items: '{{ array_details.stdout_lines }}'
	when: array_created.changed

mrlesmithjr / ansible-mdadm Goto Github PK

ansible-mdadm's Introduction

mrlesmithjr.mdadm

Requirements

Role Variables

Dependencies

Example Playbook

License

Author Information

ansible-mdadm's People

Contributors

Stargazers

Watchers

Forkers

ansible-mdadm's Issues

Recommend Projects

Recommend Topics

Recommend Org