You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We have a user who has an application-scoped bean that does lots of expensive, IO-intensive work in its constructor. This is not a good thing to do under any circumstances, but Weld could be a little nicer here.
Specifically, Weld currently uses the synchronized keyword to ensure that (as you would expect) one and only one instance of the application-scoped bean is ever created. This is not virtual thread-friendly.
If it used Lock instead, then it would be virtual thread-friendly. WELD-2788 represents a probe to see if they'll be receptive to a PR that does this.
To reiterate, the proper thing is for the construction of a bean NOT to do lots of expensive IO.
The text was updated successfully, but these errors were encountered:
(This is not a Helidon issue, but more of a tracking issue to keep track of Weld's JIRA progress.)
See WELD-2788.
We have a user who has an application-scoped bean that does lots of expensive, IO-intensive work in its constructor. This is not a good thing to do under any circumstances, but Weld could be a little nicer here.
Specifically, Weld currently uses the
synchronized
keyword to ensure that (as you would expect) one and only one instance of the application-scoped bean is ever created. This is not virtual thread-friendly.If it used
Lock
instead, then it would be virtual thread-friendly. WELD-2788 represents a probe to see if they'll be receptive to a PR that does this.To reiterate, the proper thing is for the construction of a bean NOT to do lots of expensive IO.
The text was updated successfully, but these errors were encountered: