[GitHub] [cloudstack-primate] PaulAngus commented on issue #835: [FEATURE] server: update template to another template type

2020-12-07 Thread GitBox


PaulAngus commented on issue #835:
URL: 
https://github.com/apache/cloudstack-primate/issues/835#issuecomment-739914243


   As far as I'm concerned, a 'feature' was added to the UI when in a code 
freeze. 
   
   We're deliberately exposing through the UI, a backend change that IMO wasn't 
fully thought through (as detailed above) and was hardly tested.  Therefore, I 
don't think it should ever have been merged.
   
   The determination that the admin should know what they're doing doesn't hold 
water, because they are very unlikely to know the backend mechanisms of making 
template changes. and even if they did know, CloudStack definitely doesn't do 
what it should do in some cases, and there has been no defensive testing to 
ensure it does in the rest.
   
   But I can't be bothered to argue any more. 
   
   So merge it if you want.
   
   @DaanHoogland @weizhouapache @ravening



This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org




[GitHub] [cloudstack-primate] PaulAngus commented on issue #835: [FEATURE] server: update template to another template type

2020-12-07 Thread GitBox


PaulAngus commented on issue #835:
URL: 
https://github.com/apache/cloudstack-primate/issues/835#issuecomment-739807903


   I'm afraid that that is not accurate;
   
   You're assuming that the admin knows how the backend database stores 
template data, including system VM templates. 
   Also, if an admin converts a 'user' template to a built-in template, an 
admin can fully expect that **CloudStack**  knows what it is doing and handles 
all back-end changes, in the DB and physical ones such as copying templates to 
every sec store in every zone.
   
   Also, I see no tests to ensure that if a builtin or system template is 
created (which would therefore be in all zones) and is then changed to a user 
template, that template deletion and GC occurs properly.
   
   I really hope that I don't have to go on listing functional requirements and 
possible failure cases that would need to be covered..
   .



This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org




[GitHub] [cloudstack-primate] PaulAngus commented on issue #835: [FEATURE] server: update template to another template type

2020-12-06 Thread GitBox


PaulAngus commented on issue #835:
URL: 
https://github.com/apache/cloudstack-primate/issues/835#issuecomment-739556047


   Thanks for clarifying @ravening, but that make might make it worse as:
   
   - The description does not suggest that the system VM template that was 
active is updated to inactive. - leaving the DB in an inconsistent state.
   - The testing description only says that only user -> system was tested.  
   - And I can't _see_ anything that looks like the backend operations like 
downloading system VM templates to all secondary storage pools. 
   - User VM templates are only supposed to be in one pool, so what happens 
there Cloudstack isn't going to be expecting user templates in all sec storage 
pools - hows deletion and GC going to work?
   - Builtin templates should be in all secondary storage pools as well. do 
they get downloaded?
   - Builtin templates also should be in all storage pools.
   
   



This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org




[GitHub] [cloudstack-primate] PaulAngus commented on issue #835: [FEATURE] server: update template to another template type

2020-12-06 Thread GitBox


PaulAngus commented on issue #835:
URL: 
https://github.com/apache/cloudstack-primate/issues/835#issuecomment-739542026


   The proposed UI change does not match the functionality described in the 
Backend PR which only allows changes between user and system. the UI screenshot 
shows changes to/from all types.
   
   
![image](https://user-images.githubusercontent.com/4810220/101288528-0aa18100-37ef-11eb-8dd5-dc853702aae9.png)
   
   The UI change needs reverting -admins can use the API.
   
   The [(https://github.com/apache/cloudstack/pull/3945)]  logic also looks to 
be flawed, as the description does not suggest that the system VM template that 
was active is updated to inactive.  - leaving the DB in an inconsistent state. 
   
   it **_really_** needs fixing or reverting.
   
   @rhtyd @DaanHoogland @weizhouapache @ravening 
   
   



This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org




[GitHub] [cloudstack-primate] PaulAngus commented on issue #835: [FEATURE] server: update template to another template type

2020-12-02 Thread GitBox


PaulAngus commented on issue #835:
URL: 
https://github.com/apache/cloudstack-primate/issues/835#issuecomment-737104971


   We were in a code freeze yesterday.  please revert this commit



This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org