Others have already suggested that if the schema (library) was already
created using CREATE SCHEMA then new tables created in it would be
automatially journalled.
STRJRNLIB is one way of changing an existing schema to automatically
journal new tables. Another way is to create the data area (properly)
QDFTJRN. This is what CREATE SCHEMA does. STRJRNLIB is the latest rage
but the existence of that data area takes priority. For example, if
STRJRNLIB says to journal to one receiver and QDFTJRN says another it will
journal to the one in QDFTJRN. For details on creating that data area
outside of CREATE SCHEMA see:
http://www-01.ibm.com/support/knowledgecenter/ssw_ibm_i_72/rzaki/rzakiautostrjrnl.htm?lang=en
It was also pointed out to you that journals can be set up to be "system
managed" meaning that once that receiver is
- no longer attached,
- and has been saved,
it will be deleted. By what, I'm not exactly sure. I suspect not the
save itself as there is a section on how the system retries deleting the
receiver every 10 minutes or until certain conditions are met and these
delays would not be appreciated by operators.
You can journal scrape by using some sql functions such as
QSYS2.DISPLAY_JOURNAL
See the following:
http://www-01.ibm.com/support/knowledgecenter/ssw_ibm_i_72/rzaki/rzakikickoff.htm?lang=en
We keep extensive journals,
with both full before/after images,
we even journal open/close operations,
and we keep our receivers for 80 days.
We let Mimix maintain this. We do the 80 days because we do some
extensive scraping.
We are well loved by IBM and our BP for this amount of disk.
PRTDSKINF RPTTYPE(*LIB)
% of Size in
Library Owner Disk 1000 bytes
#MXJRN MIMIXOWN 40.50 2219455512.6
ERPLXF SSA 7.96 436015644.7
QGPL QSYS 2.71 148463251.5
ERPLXSAVF SSA 2.34 128084774.9
QPFRDATA QSYS 1.78 97505812.5
ERPLXUSRF SSA 1.11 61090816.0
...
Yep, 2.2TB for journal receivers. And only 0.4TB for ERP.
Rob Berendt
midrange.com
