Question about FailOverDataConnector

[Wessel, Keith]

Thanks, Scott. I thought that was the case, and your answer gave me the clue I needed.

The attribute being used by this data connector is exported directly and used in an attribute registry definition. I also had an exportAttributes parameter on the fail-over connector which was causing it to be called directly. I removed it, and it's now only resolving from the fail-over connector when the primary fails.

It was just a learning curve for me with the move from the attribute resolver where I would have had a very obvious <InputDataConnector> triggering the dependency.

Thanks,
Keith


-----Original Message-----
From: Cantor, Scott <cantor.2 at osu.edu> 
Sent: Monday, January 29, 2024 11:08 AM
To: Shib Users <users at shibboleth.net>
Cc: Wessel, Keith <kwessel at illinois.edu>
Subject: Re: Question about FailOverDataConnector

> Is there something special I need to do to my fail over data connector to
> tell it not to run unless it's needed because the primary one fails?

Something else is depending on it directly if it's running. Contrary to what people usually think, the resolver doesn't run "everything". It assumes it should resolve all attribute definitions, and then it runs anything those attribute definitions depend on.

At least that's how it's always historically worked.

-- Scott