Thanks, Quentin.
Then it is not the log setting. When I try to trouble-shoot in a new
connection (isql rather than perl DBI over ODBC), I observe very odd results
indeed:
SQL> sparql select count(*) from <http://id.nlm.nih.gov/mesh/updates> where {
?s a ?o };
…
4393
SQL> sparql clear graph <http://id.nlm.nih.gov/mesh/updates>;
…
SQL> sparql select count(*) from <http://id.nlm.nih.gov/mesh/updates> where {
?s a ?o };
…
4
If no delete has ever been done, the number of subjects remaining should almost
2 million – 1994719.
From: Quentin [mailto:quentin@guidinghand.solutions]
Sent: Tuesday, February 21, 2017 1:41 PM
To: virtuoso-users@lists.sourceforge.net
Subject: Re: [Virtuoso-users] Scope of log_enable and effect on deletes
From the manual:
>While log on and off setting alone is reset at the end of the transaction so
>that one does not end up with logging disabled by accident, the row-by-row
>autocommit mode causes the setting to be permanent inside the calling
>connection or web request. That is, for a SQL client the setting stays in
>effect until changed or disconnected and for a web request it stays in effect
>until the request is completed.
http://docs.openlinksw.com/virtuoso/fn_log_enable/
If you want to determine whether the log is still disabled at a point in time
then this is one round-about way of doing it:
>Attempt to disable transaction log that is disabled already result in an
>error. If the second parameter is passed and it is not equal to zero then the
>error is suppressed (and it's still possible to turn on or off autocommit).
Set an exception handler to catch that error and set a flag to indicate that
the log was previously disabled. If there's no error/flag set then proceed to
re-enable the log.
There might be better ways of doing it but that should work.
--
Regards,
Quentin Serrurier.
GuidingHand.Solutions
On 2017-02-22 02:01, Davis, Daniel (NIH/NLM) [C] wrote:
I am not sure I understand the scope of the log_enable() procedure. If I
call log_enable(2,1); in a stored procedure, I am sure that it will affect the
entire connection, but, will it affect other connections as well. Would it,
for instance, persist until reboot and affect other connections? How can I
verify this – is there any command that will output the current transaction
isolation level?
Also, I have an inexplicable situation where my stored procedure loads a
dataset in NTriples format into a graph, and then does some processing on that
dataset. It is important for the processing that the graph be empty at the
beginning, e.g. before the load. Try as I might, I seem to end-up in
situations where the graph is not-empty, and I believe that this may be some
interaction between log_enable() and the delete statement. I would think
naively that log_enable(2,1); would have no effect on the serialization of
row-level operations *in the same connection*, but perhaps I misunderstand.
Dan Davis, Systems/Applications Architect (Contractor),
Office of Computer and Communications Systems,
National Library of Medicine, NIH
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Virtuoso-users mailing list
Virtuoso-users@lists.sourceforge.net<mailto:Virtuoso-users@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/virtuoso-users
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Virtuoso-users mailing list
Virtuoso-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/virtuoso-users
