listRouter provides "FAULT" redundantstate for VPC with no instance running

[sbrueseke]

problem

When all instances are stopped in a VPC with redundant routers the redundantstate will change to FAULT. This is a misleading state, because the routers are still running and working correctly. If you start an instance the routers will change to PRIMARY and BACKUP redundantstate.

versions

The versions of ACS: 4.19.2.0

The steps to reproduce the bug

1. create a vpc
2. create networks and instances in this vpc
3. stop all instances in vpc
4. use api call listRouters or take aloook at the UI in Infrastructure -> Virtual routers
   ...

What to do about it?

I would not expect FAULT as Redundant state when all instnaces are stopped in a vpc, because this indicates that something is wrong with the it. But this is not the case, after starting at least 1 instance the Redundant state gets back to PRIMARY and BACKUP.

[DaanHoogland]

@weizhouapache is this the same as #10281 ?

[sbrueseke]

@weizhouapache is this the same as #10281 ?

I do not think that it is the same. The vpc is working correctly when at least 1 instance is running. It just shows FAULT when no instance is running, but it will switch back to PRIMARY/BACKUP when starting at least 1 instance again.

So it is not really a bug, just the value redundantstate should not be FAULT when no vm is running in a vpc.

[weizhouapache]

@weizhouapache is this the same as #10281 ?

I do not think that it is the same. The vpc is working correctly when at least 1 instance is running. It just shows FAULT when no instance is running, but it will switch back to PRIMARY/BACKUP when starting at least 1 instance again.

So it is not really a bug, just the value redundantstate should not be FAULT when no vm is running in a vpc.

agree.
similar to comment #10281 (comment)

@DaanHoogland
this is a minor bug, let's add to 4.19.3 milestone

[sbrueseke]

@weizhouapache is this the same as #10281 ?

I do not think that it is the same. The vpc is working correctly when at least 1 instance is running. It just shows FAULT when no instance is running, but it will switch back to PRIMARY/BACKUP when starting at least 1 instance again.
So it is not really a bug, just the value redundantstate should not be FAULT when no vm is running in a vpc.

agree. similar to comment #10281 (comment)

@DaanHoogland this is a minor bug, let's add to 4.19.3 milestone

Ok, sorry. I did not read the comments, only the bug report. It is similar.