Update docs for changes to Recovery Service#461
Conversation
murwall
left a comment
murwall
left a comment
There was a problem hiding this comment.
Overall I think this looks very good.
A few things popped up while reading the disaster-recovery.md.
First, we have renamed the service from "disaster recovery" to "recovery service" and it would make sense to refer to the same name here.
The other thought I had is related to this text:
Provided snapshots are available, you can restore a server or a single volume to any of those snapshots. For instance, you may discover that due to faulty application logic or simply a bug, you are now experiencing data corruption. Then, one of your options would be to [go back in time](https://github.com/colder-is-better/cleura-docs/blob/78acf236ceadd536d315c491bf11eae41932d486/docs/howto/openstack/nova/restore-srv-to-snap.md) by restoring one of the available snapshots and keep going from there.
With the new recovery service, the best benefit is the possibility to restore any snapshot to a new volume. This will be the default way of restoring data in most cases. You create a new volume, attach it to your server and copy over the data that you want to restore or replace the entire mount with the new restored volume.
As for restoring onto the same volume, there are some rules that apply:
- Only the last snapshot can be restored
- The existing volume that you want to replace with a the latest snapshot, must be in status=available. This effectively means that right now you can't restore onto a volume that is used as a boot device. The only exception would be if you delete the server, keep the volume, restore onto the volume, and boot a new server from the volume. But that is just cumbersome and it would be a lot easier to just restore to a new volume and boot a new volume from it.
|
|
||
| Had you chosen to restore an **older** snapshot, then you could also restore it to a **new** volume. | ||
|
|
||
| The only time you can restore a snapshot to an existing volume is when the snapshot comes from a non-boot volume and is the newest. |
There was a problem hiding this comment.
@murwall Re this comment:
As for restoring onto the same volume, there are some rules that apply:
- Only the last snapshot can be restored
- The existing volume that you want to replace with a the latest snapshot, must be in status=available. This effectively means that right now you can't restore onto a volume that is used as a boot device.
Isn't that what this sentence here says?
* Introduce and explain new recovery service features (retention days, immutable snapshots), * make sure the old "disaster recovery" term is not used anymore.
Re-create new server guide screenshots due to changes in the CCMP UI.
Introduce and explain the new recovery service features. While at it, make sure the old "disaster recovery" term is not used anymore.
* Mention the upgrade packages during boot option, * make the note against using the console visible to both the CCMP and the CLI part of the guide, * improve syntax and brevity, and * deal with white space issues.
colder-is-better
force-pushed
the
rec-service-refresh
branch
from
78acf23 to
5eff4e3
Compare
There was a problem hiding this comment.
@colder-is-better, fix errors that are self-evident from locally running tox, please.
Due to the features of the new recovery service, rewrite the How-To guide and re-create the screenshots. While at it, make sure the old "disaster recovery" term is not used anymore.
colder-is-better
force-pushed
the
rec-service-refresh
branch
from
5eff4e3 to
6840ea7
Compare
|
@colder-is-better Again, get in the habit of running the |
This requires a new redirect_maps mapping in mkdocs.yml, so old links continue to work. While at it, we update the menu system for background documentation to now have a "Recovery service" link.
Make sure the old "disaster recovery" term is not used anymore.
In the guide for converting a boot-from-image server to a boot-from-volume one, we now use the "recovery service" term. While at it, we fix an internal link to point to the correct background document on the new recovery service.
colder-is-better
force-pushed
the
rec-service-refresh
branch
from
6840ea7 to
4357683
Compare
|
@fghaas All screenshots in the "Converting a boot-from-image server to boot-from-volume" guide should be re-created: the old "Disaster recovery" option is still visible, and there are also UI changes. Should I update the current PR or create a new one? |
|
@colder-is-better I think we've addressed all the comments that @murwall made. I'll merge and deploy. |
3 participants
Refresh public documentation in light of new recovery service features (snapshot retention days, immutable snapshots).