Why catch vagrant errors in recover?

[Darragh Bailey]

Hi,


Recently did some work on ensuring that vagrant-libvirt would respect the --no-destroy-on-error option during failure and noticed something odd about the errors that would be caught.

Looking at https://github.com/pradels/vagrant-libvirt/blob/e21465771d1bb2d2f544b6dbcf3838304ef45295/lib/vagrant-libvirt/action/wait_till_up.rb#L82 seems like no matter what error occurs once there is one internal to vagrant in any plugin or vagrant itself, it'll invariably be subclassed from the base vagrant error class and result in the recover function doing nothing.

The only other place I've come across this is within the virtualbox import action:
https://github.com/mitchellh/vagrant/blob/5ea24e39d021875c2d6cc3830236364f42f78f9f/plugins/providers/virtualbox/action/import.rb#L79


Looking around for other recover calls I see the following for the plugins command install_gem action:
https://github.com/mitchellh/vagrant/blob/41381792071413c7ed08ee55fc024e6f10b8b559/plugins/commands/plugin/action/install_gem.rb#L69-L79

Which I notice, also uses an action to perform it's function and does not check to see if a vagrant error has been defined.

What I'm left wondering is whether the code that was copied into vagrant-libvirt was unintentionally incorrect? and we should instead skip checking for a vagrant error?


--
Darragh

[Alvaro Miranda Aguilera]

Hello,

You can open an issue on the libvirt plugin repo, I am sure they will reply,

I am not sure if they monitor this mailing list.

Alvaro.

--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
 
GitHub Issues: https://github.com/mitchellh/vagrant/issues
IRC: #vagrant on Freenode
---
You received this message because you are subscribed to the Google Groups "Vagrant" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vagrant-up+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/vagrant-up/cd5a66a7-1fe6-422b-aea3-ed1033d4810f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Darragh Bailey]

On Wednesday, March 2, 2016 at 11:55:13 PM UTC, Alvaro Miranda Aguilera wrote:

Hello,

You can open an issue on the libvirt plugin repo, I am sure they will reply,

I am not sure if they monitor this mailing list.

Already talking with them, more looking for a response from a 'vagrant' perspective here rather than the plugin author.
 

Alvaro.

On Thu, Mar 3, 2016 at 4:13 AM, Darragh Bailey <daragh...@gmail.com> wrote:

Hi,


Recently did some work on ensuring that vagrant-libvirt would respect the --no-destroy-on-error option during failure and noticed something odd about the errors that would be caught.

Looking at https://github.com/pradels/vagrant-libvirt/blob/e21465771d1bb2d2f544b6dbcf3838304ef45295/lib/vagrant-libvirt/action/wait_till_up.rb#L82 seems like no matter what error occurs once there is one internal to vagrant in any plugin or vagrant itself, it'll invariably be subclassed from the base vagrant error class and result in the recover function doing nothing.

The only other place I've come across this is within the virtualbox import action:
https://github.com/mitchellh/vagrant/blob/5ea24e39d021875c2d6cc3830236364f42f78f9f/plugins/providers/virtualbox/action/import.rb#L79


Looking around for other recover calls I see the following for the plugins command install_gem action:
https://github.com/mitchellh/vagrant/blob/41381792071413c7ed08ee55fc024e6f10b8b559/plugins/commands/plugin/action/install_gem.rb#L69-L79

Which I notice, also uses an action to perform it's function and does not check to see if a vagrant error has been defined.

What I'm left wondering is whether the code that was copied into vagrant-libvirt was unintentionally incorrect? and we should instead skip checking for a vagrant error?

Possibly I didn't ask this particularly well.

What I'm really looking for is information from the 'vagrant' perspective as to how the 'recover' methods should be defined, and what exceptions should be caught by them in general. It seems like only some instances are trying to catch the vagrant error exception and there isn't a good explanation in the git log history as to why.

--
Regards
Darragh