yahoo-eng-team team mailing list archive

[George Shuklin <1673818@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

Current documentation states that:
http://cloudinit.readthedocs.io/en/latest/topics/datasources/configdrive.html#version-2

... a config drive: ...Must be a un-partitioned block device (/dev/vdb,
not /dev/vdb1)...

This is not correct.

1. Cloud-init actually, works with ConfigDrive as partition (e.g. /dev/sda1)
2. Ironic uses partition at the end of disk to write metadata, and it's absurd for baremetal provisioning to dedicate whole disk (actual SATA, SSD, SAS/FC drive) just to tiny metada.
3. According to @smoser at #cloud-init IRC, " i'm pretty sure the doc is just wrong, ... i'm pretty sure reading current code that if the filesystem has a label of 'config-2', then it will work".

I think this part of documentation should be rewritten to avoid
confusion with Ironic workflow for ConfigDrive.

** Affects: cloud-init
     Importance: Undecided
         Status: New

** Description changed:

- 
  Current documentation states that:
  http://cloudinit.readthedocs.io/en/latest/topics/datasources/configdrive.html#version-2
  
  ... a config drive: ...Must be a un-partitioned block device (/dev/vdb,
  not /dev/vdb1)...
  
  This is not correct.
  
- 1. Cloud-init actually, works with ConfigDrive as partition.
+ 1. Cloud-init actually, works with ConfigDrive as partition (e.g. /dev/sda1)
  2. Ironic uses partition at the end of disk to write metadata, and it's absurd for baremetal provisioning to dedicate whole disk (actual SATA, SSD, SAS/FC drive) just to tiny metada.
  3. According to @smoser at #cloud-init IRC, " i'm pretty sure the doc is just wrong, ... i'm pretty sure reading current code that if the filesystem has a label of 'config-2', then it will work".
  
  I think this part of documentation should be rewritten to avoid
  confusion with Ironic workflow for ConfigDrive.

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to cloud-init.
https://bugs.launchpad.net/bugs/1673818

Title:
  Misleading requirements for 'unpartitioned disks' for ConfigDrive in
  documentation

Status in cloud-init:
  New

Bug description:
  Current documentation states that:
  http://cloudinit.readthedocs.io/en/latest/topics/datasources/configdrive.html#version-2

  ... a config drive: ...Must be a un-partitioned block device
  (/dev/vdb, not /dev/vdb1)...

  This is not correct.

  1. Cloud-init actually, works with ConfigDrive as partition (e.g. /dev/sda1)
  2. Ironic uses partition at the end of disk to write metadata, and it's absurd for baremetal provisioning to dedicate whole disk (actual SATA, SSD, SAS/FC drive) just to tiny metada.
  3. According to @smoser at #cloud-init IRC, " i'm pretty sure the doc is just wrong, ... i'm pretty sure reading current code that if the filesystem has a label of 'config-2', then it will work".

  I think this part of documentation should be rewritten to avoid
  confusion with Ironic workflow for ConfigDrive.

To manage notifications about this bug go to:
https://bugs.launchpad.net/cloud-init/+bug/1673818/+subscriptions