Account for tagging the root volume after spot loss

[tewfik-ghariani]

Once a spot instance is lost or a spot request in cancelled, the root volume re-created is not appended to the block device mapping because the 'first_boot' was set to False. With this update, we are sure that all mapped volumes are consistently tagged Other

• Tag using root vol Name as depl name instead of description
• Deletion of the condition to tag the volumes
• Use of generic 'update_tags' method
• Removal of first_boot ec2 attribute

Note: The branch is based on https://github.com/NixOS/nixops-aws/pull/101 so the changes can be confusing. You can view them in this single commit : https://github.com/NixOS/nixops-aws/commit/3fcecd451e2d08cbc76cc7b2e941cc22f4061215

[tewfik-ghariani]

Well I'm second thinking this after reading the reason behind 'first_boot' : https://github.com/nixos/nixops/commit/ee60f045c4ddf006b8ba575723b1bee21279a6e4

cc @AmineChikhaoui Do you think that would break something and that I should follow a different approach?