[
https://issues.apache.org/jira/browse/FINERACT-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17089032#comment-17089032
]
Michael Vorburger edited comment on FINERACT-902 at 4/21/20, 8:13 PM:
----------------------------------------------------------------------
[~rrpawar] thanks for filing this issue, and elaborating with background on the
dev mailing list thread "Re: Re-opening Pull Request and Writing Test Cases for
[https://github.com/apache/fineract/pull/527]";.
I understand this problem (and for me it brings back nostalgic memories when
eons ago [I professionally used
OpenJPA|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.users+page:1+state:facets],
engaged [on that projects dev
list|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.dev+page:1+state:facets],
and [even contributed to
it|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.commits+page:1+state:facets]).
While changing openjpa.jdbc.EagerFetchMode to none in persistence.xml, as
originally proposed (hidden deep inside)
[https://github.com/apache/fineract/pull/738,] and then (as requested isolated)
in [https://github.com/apache/fineract/pull/777], solves this problem, it's not
the right way to address this.
Changing EagerFetchMode in persistence.xml globally very likely would have have
a very significant impact on Fineract's performance overall - it would make us
"lazy load" everything; I'd expect that to lead to a very important increase in
the numbers of SQL queries issued for all operations.
What we need to do instead is to do this much more fine grained, and use
[https://ci.apache.org/projects/openjpa/trunk/docbook/manual.html#jpa_overview_meta_fetch.]
Would you like to explore that?
was (Author: vorburger):
[~rrpawar] thanks for filing this issue, and elaborating with background on the
dev mailing list thread "Re: Re-opening Pull Request and Writing Test Cases for
[https://github.com/apache/fineract/pull/527]";.
I understand this problem (and for me it brings back nostalgic memories when
eons ago [I professionally used
OpenJPA|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.users+page:1+state:facets],
engaged [on that projects dev
list|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.dev+page:1+state:facets],
and [even contributed to
it|https://markmail.org/search/?q=vorburger#query:vorburger%20list%3Aorg.apache.openjpa.commits+page:1+state:facets]).
While changing openjpa.jdbc.EagerFetchMode to none in persistence.xml, as
originally proposed (hidden deep inside)
[https://github.com/apache/fineract/pull/738,] and then (as requested isolated)
in [https://github.com/apache/fineract/pull/777], solves this problem, it's the
right way.
Changing EagerFetchMode in persistence.xml globally very likely would have have
a very significant impact on Fineract's performance - it would make us "lazy
load" everything; I'd expect that to lead to a very important change in numbers
of SQL queries issued for all operations.
What we need to do instead is to do this much more fine grained, and use
[https://ci.apache.org/projects/openjpa/trunk/docbook/manual.html#jpa_overview_meta_fetch.]
Would you like to explore that?
> Fix error "MySQL can only use 61 tables in a join hibernate"
> ------------------------------------------------------------
>
> Key: FINERACT-902
> URL: https://issues.apache.org/jira/browse/FINERACT-902
> Project: Apache Fineract
> Issue Type: Bug
> Reporter: Rahul Pawar
> Priority: Major
>
> Error: "MySQL can only use 61 tables in a join hibernate" in PR
> [https://github.com/apache/fineract/pull/738].
> This error occurred due to two new tables - glim_accounts and gsim_accounts
> and they had Foreign Key with Gigantic tables Loan and Savings account which
> caused this exception.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
