spurious org.springframework.beans.factory.NoSuchBeanDefinitionException

[Jim Toth]

Here, we were getting errors kind of like that[1] that turned out to be due
to the jdk packages having updated, and neither the packages' built-in
scripts nor the weekly update script for non-kernel updates were restarting
or rebooting Tomcat.

It was when we were reloading on purpose, so it was immediately obvious
when it failed, which helped.

[1] Kind of, not really, but still bean related. This might have been
Tomcat rather than JDK:
[net.shibboleth.utilities.java.support.service.AbstractReloadableService:231]
- Service 'shibboleth.AttributeResolverService': Reload for
shibboleth.AttributeResolverService failed
net.shibboleth.utilities.java.support.service.ServiceException:
org.springframework.beans.factory.BeanCreationException: Error creating
bean with name 'eduPersonOrgUnitDNCustom': Cannot create inner bean '(inner
bean)#35d3a868' of type
[net.shibboleth.ext.spring.factory.EvaluableScriptFactoryBean] while
setting bean property 'script'; nested exception is
org.springframework.beans.factory.BeanCreationException: Error creating
bean with name '(inner bean)#35d3a868': Invocation of init method failed;
nested exception is
net.shibboleth.utilities.java.support.logic.ConstraintViolationException:
No scripting engine associated with scripting language javascript


--
Jim Toth (he/him)
jjtoth at vcu.edu
Lead Engineer, Middleware
VCU Technology Services


On Thu, Sep 14, 2023 at 8:40 AM Cantor, Scott via users <
users at shibboleth.net> wrote:

> > Given I flagrantly disregard the recommendation to upgrade in place ;),
>
> The fact that it's happening enough to notice when it's never happened to
> me in a single instance that didn't involve the service reloading stuff (*)
> is basically virtual proof that an older file has to be there.
>
> Otherwise it would mean Spring were completely broken, whereas the Occam's
> Razor explanation fully explains it.
>
> I guess I would start by checking for anything in the old system/ folder
> we used to create.
>
> -- Scott
>
> (*) Spring has a bug involving the use of parent beans across contexts
> when certain conditions are met, but you have to "create it twice" to see
> it, and the flows only instantiate once, and there are other caveats unmet
> in this case.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://shibboleth.net/pipermail/users/attachments/20230914/8a8aed50/attachment.htm>