>I am reposting that patch again - let me know if there are any
>questions or
>concerns.
Costin is working on a huge refactory of jk, named native2,
and in that version the jk_init is removed from child_init.
Having it in child_init was only usefull when using ajp14
to grab URIs to be handled
I am reposting that patch again - let me know if there are any questions or
concerns.
Seems like the latest j-t-c mod_jk calls initialization (init_jk) twice -
once during post_config and then in child_init. This has a side effect of
trying to create non existing workers and subsequently failing.
Seems like the latest j-t-c mod_jk calls initialization (init_jk) twice -
once during post_config and then in child_init. This has a side effect of
trying to create non existing workers and subsequently failing. Enclosed
simple patch prevents second initialization call.
Julius
Index: mod_jk.c
==