Re: Interconnection Delay
- From: Jan Burse <janburse@xxxxxxxxxxx>
- Date: Fri, 17 Apr 2009 00:36:53 +0200
joe.no_junk@xxxxxxxxx schrieb:
Are c1 and c2 used in the same thread, guaranteeing
that c1's calls are returned before c2 is used?
Are these JDBC connections? Are they in the default
autoCommit(true) mode. If so, there is no reason/excuse
for a query from c2 to miss anything c1 has done.
Joe
If c1 and c2 were pointing to different databases
which are synchronized by replication, than this
would be a reason/excuse.
http://blog.mobocracy.net/2008/09/mysql-multimaster-replication-in.html
But I observed delays also for a single database
without replication. The effect is reproducable
with JET and a .mdb database.
A reason/excuse could be the particular transaction
memory model that JET used. Wikipedia says:
... Until the transaction is committed, changes are
made only in memory and not actually written
to disk. ...
... With ODBC's in-memory policy, transactions also
allow for many updates to a record to occur
entirely within memory, with only one expensive
disk write at the end. ...
http://en.wikipedia.org/wiki/Microsoft_Jet_Database_Engine#Transaction_processing
Could it be that I was using a JDBC-ODBC connection?
And the memory policy was on?
Do other database systems also have this feature?
Maybe in a much smaler time scale than 100ms?
Bye
.
- References:
- Interconnection Delay
- From: Jan Burse
- Re: Interconnection Delay
- From: joe.no_junk@xxxxxxxxx
- Interconnection Delay
- Prev by Date: Re: Interconnection Delay
- Next by Date: Bill Of Material UML and Hibernate file
- Previous by thread: Re: Interconnection Delay
- Next by thread: Bill Of Material UML and Hibernate file
- Index(es):
Relevant Pages
|