Eager Fetch Mode limit clarification

Question

Hi,
I would like clarification of the following from the 3.3.4 docs
(kodo-jdo-3.3.4/docs/ref_guide_perfpack_eager.html):
"Once Kodo eager-joins into a class, it cannot issue any further eager
to-many joins or parallel selects from that class in the same query.
To-one joins, however, can recurse to any level."
Given the following class/table relationships:
X --> Y -->> Z
(where --> is one-to-one and -->> is one-to-many)
Does this mean that if I query X, Y can be eagerly fetched, but not Z?
This is certainly what we see - a join to fetch X and Y in one select
statement and then many one-off selects from Z for each Y.
Is there any way to load Z more efficiently? It would seem that once
Kodo has loaded the Y's that it knows all it needs to load the Z's with
one SQL select - Z would have a FK to Y, so "select ... from z where
y_id in (<kodo puts the loaded y ids here>)" would seem feasible,
especially given Kodo does this in other places.
Thanks,
Richard

Answer 1

Thanks Abe.
Abe White wrote:
X --> Y -->> Z
(where --> is one-to-one and -->> is one-to-many)
Does this mean that if I query X, Y can be eagerly fetched, but not Z?Yes, unfortunately.Our actual case also includes the following relationships: X -->> W -->>
V and strangely, they do seem to be be eagerly loaded when we query for
X. i.e. one query to load all Vs. How does this work given the above?
Is there any way to load Z more efficiently?Typical workarounds include querying for Y's instead of X's and using
the inverse relation from Y to X, or if there is no such relation,
performing 2 queries: one for Y's and one for X's.Are there any plans to support greater depth eager loadings in any
upcoming Kodo version?
Thanks again for your help,
Richard